Advancements in Smart City and Intelligent Building

Due to the difficulty of establishing the accurate control model for building an intelligent system, a neural network predictive control method is proposed, in this paper, based on a weed optimization algorithm. Through considering indoor temperature and relative humidity environment factors, a control model of temperature and humidity environment is first established in an intelligent building. Then, the hidden layer nodes center of the RBF neural network is optimized by using the weed optimization algorithm. The above mentioned work focuses on improving the shortcomings of Orthogonal Least Squares (OLS) algorithm, and simultaneously simplifies the network architecture. The simulation results show that the RBF neural network predictive control method based on the weed optimization algorithm has better approximation ability and generalization ability contrasting with the OLS algorithm.

Currently, the demand on thermal comfort of architectural environment is becoming higher and higher, the establishment of thermal comfort model based on the physiological parameters plays an important role in improving the indoor thermal comfort and building energy efficiency. The mean skin temperature, skin conductance, and heart rate are three main physiological parameters used to characterize the thermal comfort state of the human body, which will be of great significance to establish the thermal comfort model. In this paper, the physiological experimental program was designed in detail and the subjective questionnaires ruler was classified involving thermal sensation, thermal comfort, and sweat rate. Then, the human subjective thermal response distribution and the regular pattern were analyzed according to the vote on the state of thermal comfort and physiological parameters of subjects under five experimental conditions. As a result, the multiple physiological modeling of indoor human thermal comfort was performed by using partial least squares (PLS) method based on mean skin temperature, skin conductance, and heart rate. Furthermore, the experiments aimed at evaluating the accuracy of the established indoor human thermal comfort model were performed, the results indicate the accuracy of the established model is satisfactory.

To reduce fossil fuel energy consumption and improve the using efficiency, it is of great significance to study the thermal comfort model based on multiple physiological parameters. Compared with the classical models, comfort can be reflected more accurate by using the thermal comfort model based on multiple physiological parameters. In this paper, the experiments were performed to verify the effectiveness of the thermal comfort model. In particular, to verify the practicability of the thermal comfort model based on multiple physiological parameters, the established thermal comfort model based on the genetic algorithm with a backpropagation neural network and the classical PMV were compared. The results indicate that the established thermal comfort model is reasonable, which provides a feasible option for achieving a comfortable indoor environment. Finally, it puts forward further study on the thermal comfort model based on more physiological parameters.

The combined cooling, heating, and power (CCHP) system is beneficial to energy conservation and pollution elimination. With the increased application of the distributed photovoltaic (PV) system, the CCHP and PV combined system is attracting the attention of energy users. Due to the randomness of PV power, the operation dispatch of the combined CCHP and PV system becomes even complicated. This paper proposes an optimization model for this system to achieve the minimization of the cost, by considering the seasonal change of the electric power and heating demands. An improved particle swarm optimization (PSO) is adopted in the optimization model to enhance effectiveness. The case study for an office building is also presented in this paper. And the results show more than 10% of the cost is saved.

Control and optimization of the quality of the indoor environment are necessary to ensure indoor comfort and reduce building energy consumption. Indoor environmental quality that contains a variety of uncertainties and nonlinear factors is difficult to be described by the traditional linear system. In this paper, by defining the linear relationship between physical parameters and control parameters of the indoor environmental quality, the control, and energy consumption optimization modeling is established according to the data measured based on a bilinear model. On this basis, this study proposes a model predictive control system coupled with an intelligent optimizer for indoor environmental quality control. Ant colony optimization (ACO) is utilized to optimize the building energy management. Experimental results show that the proposed intelligent control system successfully manages indoor environmental quality and energy conservation.

The data quality of building energy consumption monitoring platform is generally not high and there are a lot of problem data. This paper proposes an identification method of implicit error energy consumption data based on the overall usage characteristic. In this method, we connect hourly energy consumption data into lines. According to the influencing factors of the building operation, we classify the energy-usage mode. By using the clustering method, we identify partial abnormal data. Then we count the slopes of historical energy consumption data characteristic lines and compare the time-varying characteristic lines of real-time with the historical characteristic lines under the same energy-usage mode. By using the energy consumption data of an office building, we verify the reliability of this method in identifying the abnormal energy consumption data of lighting and socket. This method improves the quality of data and will make the energy monitoring platform more efficient in building energy conservation.

In this paper, a weight control system is developed based on Beckhoff PLC for a ZJ19 cigarette machine, which improved the performance of the existing system, and the control system is also introduced. The system uses the microwave to measure cigarette weight, and uses the digital PID control algorithm to achieve weight control. The system hardware structure is simple, the performance of the system is stable and reliable with quick response set aside the upgrade space. The original machine control system was completely replaced with the designed system to meet the requirements of slender cigarette production precision and the production efficiency and product quality were ensured.

In view of the characteristics of two-wheeled robots, such as the small size and flexible movement, and the increasing demand for environment and security, this design implements a building environment mobile monitoring and security early warning robot. On the basis of the self-balancing of two wheels, the robot can monitor the building environment and alert people with voice intelligently. The system combines automatic control technology, PWM DC motor control technology and sensor technology. It applies the STM32 as the controller and uses the PID control algorithm to meet the requirements of its characteristics, such as multiple variables, nonlinearity, strong coupling, parameter uncertainty, and others. The wireless sensing and control are also established. For this, people can use the mobile phone or PC terminal to realize the wireless mobile monitoring, early warning, and alarm of building environment including temperature, humidity, illumination, air pressure, altitude, air quality, fire source detection, etc., based on mobile machines.

To improve the efficiency of resource utilization of large public buildings, indoor positioning and navigation is more and more possible in recent years. But the navigation accuracy is poor because of the complex environment in the buildings and the failure to use the satellite navigation signal. To solve this problem, a navigation and integrated information system is designed and developed by using the existed wireless network signals. A position fingerprint location method is adopted in order to ensure the finite complexity of the algorithm. A weighted K-nearest neighbor algorithm based on probabilistic reasoning is proposed to meet the requirements of navigation precision. This system will be widely used in large public buildings such as market, museum, library, and transport hubs.

In order to improve the atmospheric environment monitoring mechanism and realize the construction of environmental maps, this paper proposes a proximity factor concentration fusion method for the problem of sub-map fusion to the construction of multiple rotorcrafts maps. The method refines the neighboring sub-maps to coincide with the boundary concentration factor and calculates the concentration factor of the overlapping area based on the factor mean algorithm to obtain a complete gas concentration map. The fusion of two sub-concentration maps is taken as an example in this paper. The two sub-concentration maps with short time difference before and after are fused into a map, and then the feasibility of the fusion method is verified by Fluent and Matlab simulation experiments, which provides the research foundation for the fusion of multiple gas concentration maps.

This paper presents an understanding of active noise control (ANC) systems especially focus on the physical geometries of the detector and the observer and develops a series of simulation experiments for different types of signals to test the cancellation performance using the feedforward active noise control procedure with the fixed controller proposed by Leitch and Tokhi (IEE Proc PS-A 134(6):525–546, 1987). This study demonstrates the effectiveness of the proposed control approach and also builds up a foundation for applying the proposed new ANC algorithms in future studies.

Recently, with the rapid development of the domestic prefabricated concrete units industry in China, the layout of the beam field plays an exceedingly significant role in the efficiency and cost of the production on the prefabricated concrete. The research builds the assessment system based on the analysis of the process flow of the prefabricated box beam and combines with some relative factors on workshops of the prefabricated concrete, which designs the layout scheme of three workshops according to the unit area. The author gets the optimal layout design scheme of the prefabricated box beam by using the module of AHP (Analytical Hierarchy Process) and yaahp software which are used to calculate and analyze three layout schemes.

Smart city, as a developing strategy for a global economy, has been initiated in many countries and presented the booming future. There are various researches suggesting that a blockchain is an indispensable tool for smart city construction, but how to adopt it to smart city strategy is not certainly solved. The study here aims to propose a scheme of employing the blockchain in smart city strategy with the theory of Knowledge Governance and Dynamic Alliance, of which the former is focusing on reducing the organization of dynamic alliance, deciding the information transmission in blockchain, and the latter is organizing the smart services by promoting the knowledge innovation and competition and quickening the smart solution with blockchain in city. The scheme is analyzed and its function is detailed at last in order to help explain how it works for guiding and planning the smart city model.

The attendance system plays a very important role in the modern enterprise’s operation, and the security of the building has always been a matter of concern to the people. Based on networked surveillance video, this paper integrates the attendance and security functions and fuses video image processing, deep learning, and face recognition to design an intelligent attendance and security system. We propose a sliding average method to identify persons’ identities. The experimental results verify the effectiveness of our method. The false reject rate (FRR) in our system reaches 0.51%, the false accept rate (FAR) reaches 2.52%, and the correct identification rate reaches 98.85%. The system is applied to some video surveillance areas, with advantages of nonintrusive, passive attendance and multiple persons’ attendance at the same time.

In order to improve the efficiency of occupant evacuation in high-rise and intelligent buildings, this paper analyzes the disadvantages of traditional emergency evacuation in high-rise buildings and conducts the feasibility analysis of using an elevator as the evacuation tool. With the help of evacuation software, Pathfinder, simulation experiments are given to study the collaborative strategies among stairs and elevators under partial and total building evacuation. Results show that with the guarantee of safety and reliability, an elevator is a very efficient tool to evacuate occupants in high-rise buildings, especially those mobility impaired occupants. Combination of stairs and elevator can not only reduce the evacuation time but also shorten the occupants’ travel distance.

Nowadays, it has been an important issue to evacuate mass occupants for safety from large public buildings under emergency conditions. However, evacuation design in public buildings still remains in a static plan, which is difficult to deal with diverse types of hazards and problems of uneven personnel distribution. Thus, a set of guidelines for safety design of large public buildings, or even a set of dynamical emergency evacuation systems, is necessary. In this paper, an emergency evacuation system for large public buildings is proposed based on intelligent building by means of two commercial simulation platforms—PyroSim and Pathfinder.

A questionnaire survey on the status of intelligent building development was carried out. Based on actual survey questionnaire data, the development status of technology and applications was described in this paper, mainly discussing the existing problems and the proper future development. It is aimed at providing valuable findings on intelligent building development, exploring an innovative way to build a practical, reliable, and easy-used intelligent system to meet the requirements of users. Among all the designing aspects, the configuration of both hardware and software is the most tedious and error-prone part when building an intelligent system, as a huge workload is inevitable when solving the difficulties on system transformation. Unsuitable operation logic of the system and equipment is another problem, as the distance between the original expectation of the intelligent system and the real operation effect is still large. Though some building systems realize integration between the subsystems, they work independently, thus enabling different subsystems to work as a whole is another problem which needs to be emphasized. A much simpler configuration process or structure of the software and hardware system is significantly required in this field. If the controller could adapt its own operation logic automatically according to preset equipment parameters, this problem might be solved fundamentally.

It can result in substantial energy saving in heating, ventilation, and air-conditioning (HVAC) system by improving the control strategy of heating, ventilation, and air-conditioning system. However, it is challenging to obtain the optimal control strategy of an HVAC system due to its model’s complexity. In this paper, a regression model is proposed for the wet-bulb temperature which is a key variable in cooling tower and fan coil unit. The proposed model avoids the iterative computing process of obtaining the value of the wet-bulb temperature and reduces the complexity of an HVAC system’s model. Numerical results show that the proposed model takes less than 7% computing time to get the value of wet-bulb temperature, and the relative deviations are less than 0.4%, compared to the original model.

The constraint-based nonlinear multivariate function optimization algorithm was used to optimize the distribution of cooling load between chillers and ice-storage tanks. The goal is to minimize the cooling load and system running costs of the air-conditioning system. Based on the peak-valley price principle of the power grid system, the most economical running of the ice-storage air-conditioning system is achieved. The results show that compared with the traditional ice-storage air-conditioning system control algorithm, the proposed method can reduce the power consumption of the system by 10.32% and reduce the system operating cost by 12.07% under the premise of satisfying the demand for terminal cooling capacity.

In order to deal with the issue of lacking universality in the centralized optimization algorithm for parallel pumps, a fully decentralized optimization algorithm based on log-linear model is proposed to complete pumps group operation, consisting of many same type pumps, under the least total power consumption. Through analyzing the characteristics of distributed control system of parallel pumps and the characteristic model of pumps, the difference between centralized optimization model and distributed optimization model of pump system is explained. Introduce the probability model based on log-linear model to calculate the probability distribution on speed ratio space of each pump at each iteration and determine the operating strategy of each pump based on probability distribution at the last iteration. Finally, take a chilled water circulation system in practical project as a study to research and certify the effectiveness of the mentioned algorithm by simulate experiment. The simulation experiment shows that the algorithm can optimize the number and speed of the same type of parallel pumps.

The high false alarm rate and the difficulty of modeling are the main problems in the field of cooling tower system fault detection which is an important energy consumption optimization method in heating, ventilation, and air-conditioning (HVAC) system. This paper proposes an effective solution that is used to reduce the false alarm rate and built a gray box model which simplified from the physical principle of a cooling tower. The Kalman filter is used to forecast the running state of the cooling tower system, and the dynamic control limit set by the statistical process control (SPC) is used to reduce the false alarm rate. Through the final experimental results in the Sino-German building, located in the northeastern part of China, it can be seen that the control limit can be effectively adjusted according to the fluctuation of the natural environment, and the false alarm rate can be well controlled.

With the rapid development of intelligent buildings and the continuous improvement of management requirements, demand for real-time information management and rationality of the storage strategy design in a central air conditioning system are increased simultaneously. Aiming at the characteristics of the central air conditioning data management system with large numbers of detecting points, large capacity of storage, variety in subsystems, and management complexity in intelligent buildings, the data acquisition module was constructed and the data management system was designed and implemented in the research based on OPC remote access server technology and the zero configuration and high efficiency of SQLite database. The experimental and practical results show that the designed system can be used for data acquisition and information management in central air conditioning monitoring systems and improve the level and efficiency of management to a certain degree.

That the aging equipment, nonlinearity, and other external factors combine to make one-order plus time-delay process model be uncertain in Heating, Ventilation, and Air Conditioning (HVAC) makes the initial control parameters ineffective. Aiming at one-order plus time-delay process model with uncertainty in HVAC, the Model-Free Adaptive Control (MFAC) is introduced in studying the control characters. The changes on the overshoot and settling time are simulated and compared with PID, which shows that MFAC has good stability and anti-interference and is insensitive to the change of time-delay. Therefore, it is proved that MFAC is suitable for solving the problem of control failure caused by changing time-delay. However, there are no good methods for the parameters setting of MFAC, which makes it difficult to find the optimal controller parameters. Aiming at the condition, the simplex method is used to optimize the controller parameters and the first-order inertia plus time-delay model is regarded as the controlled object in this paper. The simulation results show that the MFAC parameters optimized by the simplex method have good control results.

In actual operation, due to the change of many factors, the central air conditioning system runs at non-designed working conditions in most of the time. Usually, it works under partial load, cannot meet the maximum load, causes great waste of energy. This paper proposes an Iterative Learning Controller algorithm (ILC) for the air conditioning water system, deal with the variable frequency control for the secondary pump. Optimization settings of water pressure differential value are given according to customer demand and based on the water valve features, thus make water valve in the chilled system having the largest opening as far as possible to provide the required minimum water differential pressure. By this way not only good control effects can be obtained, but also the energy consumption of pump delivery can be reduced.

In this paper, the online evaluation system for a chiller plant is proposed, which is significant for improving the energy efficiency and operation security of the chiller plant. For a general evaluation result, the system can estimate the chiller plant from three aspects: operating security, system performance, and cooling-supply quality, respectively. For each aspect, the paper presents a dynamic scoring method, which can convert the abstract evaluation indexes into the visualized contents by different evaluation demands. At the end of this paper, the effectiveness of the system is verified to perform excellently in practicability and operability, and it can also be used in most of the chiller plants.

In this paper, a new steady-state power consumption model using the Elman Neural Network (ENN) is proposed. The model is dependent on the external parameters of chiller, which are easily monitored and which are related to the global optimization of an air-conditioning water system. The simulation results show that the model can complete the training process within 3 s. In addition, it can be seen that the results of the model are in good agreement with the experimental values with the majority of the RE values within ±3%. Therefore, this model is suitable for on-line prediction of the power consumption of chiller in on-field engineering.

It is a fact that comparing the effect of signal change (especially changing from green to yellow) on driving decisions in different seasons is a significant way to promote traffic safety. In this study, by using video recording, a variety of data through observing vehicles on four intersections in Shenyang were obtained to analyze drivers’ driving decisions during green countdown signal change in summer and winter. Based on data including speed of a vehicle, the remaining time of countdown timers, and the Electronic Police (E-Police), and a Logistic model was established to analyze the influence of signalized intersections on drivers’ decision-making in summer and winter. The result shows that the stop decision is only impacted when the countdown signal is coexistent with E-Police in summer. Also, both the countdown signal and the E-Police have a great influence on the stopping decision in winter.

This paper aims to evaluate the performance of taxi firms and discuss its improvement for those underperforming taxi firms. The performance measured from the operational efficiency and service effectiveness perspectives has a significant impact on the input–output balance in the taxi market and economic benefits and social welfare. Based on this, we propose Data Envelopment Analysis (DEA) method to achieve the measurement of performance and adopt the benchmarking approach to complete its improvement. In doing so, this paper first discusses how to make the appropriate selections in terms of decision-making units, DEA models, and input–output indicators. The input-oriented DEA models are then applied to assess different types of performance indicators. And moreover, to provide the practical improvement recommendations, the targets assigned by the benchmarking approach and the gaps between targets and current values are also presented and the future effort directions are thus pointed out. Our results suggest that assessing and improving the performance of taxi firms are conductive to make and carry out management strategies in the taxi market.

Localization is the foundation of automatic driving and it has always been a topic of research hotspot and difficult to deal with. The DV-Hop algorithm is most widely used in node localization research, and it will be used together with particle swarm optimization algorithm. In this paper, an NJPDH algorithm is proposed to compensate for the inferior accuracy of DV-Hop localization algorithm. Based on the DV-Hop location algorithm, the algorithm is added to the weight of each beacon node, the average distance is weighted, the particle swarm optimization is optimized from two aspects of the inertia weight and the active factor to avoid the particles trapped in the local optimization, and then get the location of the unknown nodes better. The simulation results show that under the same hardware conditions, compared with the particle swarm based PSO-DV-Hop algorithm, it can effectively reduce the impact of the jump distance, increase the coverage of the nodes, improve the positioning accuracy and robustness of the positioning process, and have better applicability.

In this paper, a new coordinated control model is proposed based on the vehicular operational features, and a multi-objective optimization algorithm NSGA-II is employed to the model for the operation of the vehicle traveling on an urban arterial road taking three evaluation indexes into consideration as the average vehicle delay, the queue length, and the vehicle exhaust emission. A numerical experiment was made in an urban arterial road with three intersections on VISSIM for the proposed strategy, and the simulation results were compared with two commonly used pre-timed methods: Webster’s method and MAXBAND coordinated control method to verify the effectiveness of the proposed strategy in dealing with the unbalanced traffic volume condition, and it proved its advantages in designing and managing traffic systems more efficiently.

Whether the intersections of the tramway are closed have a very significant influence on the operating efficiency of trams and vehicles. How to coordinate the right-of-way relationship between trams and vehicles is of vital importance improving the overall traffic efficiency of the corresponding intersection. Based on the historical data of the intersections along Hunnan Modern Tram No. 5 line, we studied the relationship between the ratio of traffic flow to non-winding traffic flow and the delay. A single-factor analysis of the applicability of closed conditions of the intersections was carried out by simulation, and the optimal applicability conditions were achieved. The conclusions of this study provide a scientific basis for intersection signal timing optimization of the tramway and traffic system.

In order to alleviate the internal traffic situation, the first- and second-tier cities have introduced traffic control measures to restrict the driving roads and period of freight vehicles, which makes the operation of distribution enterprises more difficult. Based on the recursive matrix of vehicle arrival time under soft time window, a distribution path optimization model considering urban freight vehicle traffic control is constructed, and a hybrid genetic algorithm based on “chain coding” and “sub-chain compensation” crossover operator is developed to solve the problem. The changing trend of total distribution cost under different control scenarios and the comparison of convergence quality under different sub-chain size thresholds show that the proposed model and algorithm are effective and practical.

In the NCHRP Report 672, a brief introduction is given to the right-turn slip lane at a roundabout, and the lane capacity is derived. On this basis, we considered the setting mode of two types of slip lanes, such as free driving (Free Flow, FF) and yield driving (Yield, Y). The traffic flow characteristics were analyzed under the conditions of its setting. A large number of scenes were designed by VISSIM simulation, taking the average vehicle delay as the index. A comparison of operation was conducted between with and without slip lane, under the combination of different traffic flows and right-turn ratio. The critical conditions of different setting modes were preliminarily discussed. Considering the impact of slow traffic such as pedestrians and bicycles, the next research plan was put forward.

The minimum concave cost transportation problem is the benchmark problem in numerical computing and it has been used widely in the schedule of smart transportation. In this paper, a deterministic annealing neural network algorithm is proposed to solve the minimum concave cost transportation problem. The algorithm is derived from two neural network models and Lagrange-barrier functions. The Lagrange function is used to handle linear equality constraints and the barrier function is used to force the solution to move to the global or near-global optimal solution. The computer simulations on test problem are made and the results indicate that the proposed algorithm always generates global or near global optimal solutions.

This paper analyses the behavioral characteristics of lane-changing at signalized intersections. It analyses the expected speed of a vehicle running and the benefits of lane-changing, which is based on the theory of traffic flow at first. Second, it sets up a basic model of the vehicle lane-changing in the intersection, combined with the relative speed, the relative distance and the space required for the vehicle in the lane-changing process. Finally, it analyzed the influence of vehicle lane-changing behavior on traffic flow using traffic flow data which was obtained by the actual survey. Research results show that the original lane traffic decrease by 4.6% and the target lane traffic decreases by 8.1% in 12 signal cycles, which is caused by vehicle lane-changing in the intersection.

A distributed optimal control algorithm is investigated, in this paper, to deal with the problem in a peer-to-peer (P2P) control setting for parallel pumps in heating, ventilation, and air-conditioning (HVAC) systems. Each pump is equipped with a controller and becomes an intelligent node (as a peer), and nodes are equal, self-organizing, and mutually coordinated. When the HVAC system provides the pressure difference and flow rate, each intelligent node applies the random generation of the speed ratio samples method. Then, the nodes coordinate with each other and constantly optimize the speed ratio of each pump, so that the total energy consumption of the system tends to be minimum. Simulation experiment results are provided to demonstrate the performance of the proposed algorithm.

Nowadays, distributed systems turn to be increasingly competitive with centralized systems, especially in the construction field. Thus, it is indispensable to design new effective distributed algorithms. In this work, a distributed algorithm focusing on the building space topology matching problem is proposed. In distributed control architectures, each floor is equipped with several Computing Process Nodes (CPNs) which play roles as controllers in the system. Our goal is to make every CPN in the building system acquire its own position on a designed CAD drawing. To achieve this, we utilize the geographic relationships among nodes. In the algorithm, each node compares its local topology features to the ones in the drawing and communicates with their neighborhoods. We prove that the topology matching problem can be solved by using this algorithm and derive the upper bound of the number of iterations. The experiments show that this algorithm works successfully in real designed buildings.

The optimization of a complex system with multiple subsystems is a tough problem. In this paper, a Decentralized differential evolutionary algorithm (DDEA) is proposed. The simulations for both DDEA and centralized DE on three benchmark functions are carried out. The numerical results show that DDEA is efficient to solve decentralized optimization problems. On these problems, the proposed DDEA outperforms centralized DE in convergence.

A novel fault diagnosis method is proposed in this paper in the distribution network based on wavelet transform and a Bayesian network. After the wavelet transform, decomposition, and reconstruction of various electrical basic quantities by amplitude, phase angle, and energy, the electrical fault feature quantity is combined according to various weights, and then the corresponding component switch fault characteristics are calculated by Bayesian. A simple Bayesian fusion of the electrical fault feature and component switch fault characteristics is used as the eigenvector of the Bayesian network, and then trained and predicted by Bayesian network. The experimental simulation results show that the fault diagnosis method for power distribution network based on wavelet transform and Bayesian network proposed in this paper has an obvious recognition degree according to the single fault feature. It is very accurate to identify the type and faulty components.

The supply and distribution fault location method of wavelet packet energy gray-level moment wavelet neural network based on dragonfly algorithm (DA) is proposed. First, it makes convolutional wavelet energy gray-level moment for faulted electrical information, and then extracts the energy gray-level moment of fault features as the eigenvector of wavelet neural network for training, while the network parameters of wavelet neural network are globally optimized by DA. Finally, compared with the supply and distribution network fault location algorithm of support vector machine based on the DA and the pure wavelet neural network for supply and distribution network fault location algorithm, the experimental simulation shows that the proposed method has faster convergence speed. It also has the advantages of rapid descent speed, high precision, less iterations, rapid convergence, high locating accuracy for faults, and short locating time.

The decentralized building intelligent system is a new type of building intelligent control system, which forms a decentralized network through intelligent nodes. It solves the difficulties in the configuration, modification, and information sharing of traditional centralized building control systems. However, because of the characteristics of no-center and parallelism, the decentralized system still lacks effective programming language support. In this regard, this paper designs a dedicated graphical programming language for the decentralized building intelligent system, then describes and explains from the static definition and dynamic interaction of graphics element two aspects. Finally, combining a case of control, this paper draws the connection diagram with the graphical programming language and makes an explanation, which verifies the feasibility of the graphical programming language for the decentralized system.

This paper introduces Insect Intelligent Building (I2B), a new architecture of building control systems based on Internet of Things (IoT). It has the following key features that are missing in the traditional solutions: (1) it has built-in agent models for space units and control devices; (2) it has built-in link models for immediate connections between space units and control devices; (3) it assigns a smart controller (called insect) for each space unit or each control device with six data ports (called legs) and these controllers communicate with each other only through their data ports according to the link models; (4) it decompose various building operation/management commands into computing tasks that run on smart controllers and depends on the collaborations among neighboring controllers to achieve the desired effects of the commands. Different from the traditional centralized solutions, this new architecture allows the building control systems to be developed in before the buildings are really built and simplify the installation and configuration of the building control systems.

This paper introduces the application of Relevance Vector Machine (RVM) in fault diagnosis. First, the theoretical contents of RVM including the algorithm characteristics, the derivation of mathematical models, the characteristics of kernel functions and multimode classification are introduced. Then, the multi-classification fault diagnosis model of building electrical system is established by using RVM. Finally, the experimental results show that the RVM model has good classification effect on small sample data.

This paper presents an intelligent home control system based on PLC. The system combines the BP neural network algorithm with the traditional PID algorithm. The proposed algorithm has better dynamic performance, faster adjustment speed, and smaller overshoot. It can track to control and adjust the complex and nonlinear data with a higher standard. At the same time, the overall design plan of the system is described in details. The system designed in this paper provides a good reference for the practical application of intelligent home.

Knowing the number of occupants and where they are located proves crucial in many smart home applications such as automated home control, anomaly detection and activity recognition. In this paper, we propose a novel classification-based occupant counting method that makes use of existing and prevalent WiFi probe requests that are originally designed for WiFi devices to scan WiFi APs at certain channels. First, we employ a binary-location-classification model to determine each detected occupant inside or outside a targeted area; then the neural network is introduced to act as the classifier. Moreover, multiple WiFi sniffers for each given target area are deployed to generate multiple features for the neural network to perform classification and it proves mathematically to be more accurate than one WiFi sniffer only used. Finally, we validate our proposed method through real experiments. Results show that our classification-based occupant detection method using multiple WiFi sniffers outperforms the 1-WiFi-sniffer-based method, and its accuracy makes it suffice to be a viable approach to occupant estimation for smart home.

Considering that existing device-based occupant detection methods cannot count those who do not carry a device, in this paper, for buildings where the behaviour of the occupants tends to be regular, taking the WiFi-based occupant detection method as a basis, we propose a p-persistent frequent itemsets with 1-right-hand-side (RHS)-based occupant detection algorithm to improve the occupant detection performance in terms of accuracy. Association analysis using apriori algorithm is utilized to predict the occupancy of buildings through mining the relationships among occupants. We mathematically prove the reasonability of frequent itemsets with 1-RHS chosen in our algorithm and show the experimental results of applying this approach with different p. The results show that our proposed method can improve the accuracy performance in that it can see the occupant in buildings that the WiFi-based occupant detection method cannot see.

Microgrid is an effective way to accept distributed renewable energy, and the development and application of renewable energy can effectively solve the current energy and environmental crisis. However, due to the uncontrollable and intermittent nature of renewable energy, coupled with the complexity of the operation modes of the microgrid, it is more difficult to optimize and control its operation, which has become a key issue in the energy management of microgrid. Considering the randomness of renewable energy, a multiple timescale optimization plan is proposed, which is a two-stage optimization scheme. The scheduling period of day-ahead optimization is 24 h. The targets of load supply and cost are selected as the objective function of an independent microgrid, and the power constraint of each distributed power source is set. The particle swarm optimization algorithm is used to optimize the system. Short-term optimization optimizes the results of day-ahead optimization for a second time, and takes 15 min as a scheduling period. The objective functions of revisions to the plan of day-ahead and the cost are selected and solved by the particle swarm optimization algorithm. The results are verified by the particle swarm optimization and show rationality and feasibility of the method proposed in the paper.

The central heating is a complex nonlinear system. It is difficult to establish an accurate model based on multiple heating substations. In this paper, the Long Short-Term Memory (LSTM) algorithm is proposed to solve this problem. Heating substations generate data with the time series characteristics. The algorithm not only reflects the characteristics of time sequence of heating substations, but also solves the problem of long-term dependence. And, the necessary information can be saved in a limited memory capacity. Based on a large amount of historical data of the heating system of a Baotou heating company, ensuring that the total heat source is sufficient, the simulation results of the LSTM model in multiple substations show the validity, which provides the basis for the optimization of the central heating system, and a reference for LSTM to solve the complex time series modeling and prediction problems.

In urban district heating systems, the change of heat load is greatly influenced by various exterior factors. In order to meet the demand of heating system while achieving energy conservation and environmental protection, it is in this study, many kinds of artificial neural networks are compared, and a kind of Elman neural network is proposed for modeling heat load forecasting based on the temperature and the sunlight factor. The method obtains the real-time weather temperature from the Application Programming Interface (API) interface of the meteorological web site, added the illumination intensity as an input of the heat load forecasting model, and established the sample data sequence of the forecasting model. The real-time data is used to update history data and it makes up the new inputs to achieve short-term heat load rolling forecasts. The simulation results show that this method can accurately predict the future heat load, and achieve the purpose of on-demand heating, energy conservation, and environmental protection.

In order to solve the problem of thermodynamic imbalance commonly existing in the heating system, an adjustment method that can realize the uniform heating to a building is put forward. In the single pipe system, it regulates the flow based on the energy balance between the heat of system supply for users and the heat dissipation of radiator to indoor, reversing the direction of the supply water and return water. By means of theoretical analysis, the results indicate that this method can solve the problem of building overheating, and the energy saving rate is higher than 30%. This method can also solve the problem of uneven heat and cold of the building by adjusting the ratio of positive and reverse times.

In this paper, fault diagnosis methods based on illumination variation similarity analysis for High-voltage Light Emitting Diode (HV-LED) lamp driven by segmented linear driver are proposed. The proposed methods assess the illumination variation similarity between the diagnosed lamp and the different fault-type lamp to diagnose whether the diagnosed lamp occurs fault or not and confirm the fault type. Euclidean distance is applied for calculating illumination variation similarity. The proposed fault diagnosis methods contain four parts: illumination signal smoothing, similarity calculation, similarity assessment, and fault recognition. The illumination variation data of the fault-free and three different faulty lamps, which are based on one HV-LED lamp driven by four-segment linear driver, are investigated for method verification experiments. The experimental results and analysis are given to demonstrate the validity and effectiveness of the proposed fault diagnosis methods in test chamber environment.

This paper proposes a novel approach to spatial illumination calculation by experiments. Existing researches on illumination calculations have these problems such as large errors in the calculation of optical metrics, and there are great limitations in the spatial form and layout of devices. In order to solve these problems, based on the light distribution curve and the law of Lambertian reflectance, the spatial illumination can be obtained, and the mathematical model of a complex spatial light environment with more accurate calculation results and wider application range was established and implemented by MATLAB. Finally, it is used in the illumination calculation simulation of different spatial models. After calculation, the average illumination intensity calculation of space work surface is compared with that of DIALux results. The error is within ±3%, the distribution of plane illumination curves is basically the same, and the solution to the circular bottom surface space, the unequal height of the bottom spatial light environment cannot directly describe the problem.

Inspired by the excellent work of wireless sensing, we propose a non-invasive activity recognition system, Under-Sense, for underground space sensing with a pair of commodity Wi-Fi devices. Firstly, by extracting relative phase information from all 90 subcarriers, we construct fine-grained images and then compress the rectangle images into k-dimension by singular value decomposition (SVD). A nine-layer convolutional neural network (CNN) is designed to automatically extract important features from constructed images and classify five human activities. Our results show Under-Sense could achieve 99.5% average accuracy in the empty meeting room and 96.7% in complex student studio environment.

In order to ensure the successful completion of the tasks of a system, the accuracy of data is the foundation and indemnification. To complete the data verification on the decentralized computing platform, a decentralized Kalman filter with state constraints is presented in this paper. The decentralized sensing architecture takes the form of a network of transputer-based sensor nodes, each with its own processing system. So it does not require any central processor or common clock. Based on that, this new algorithm can allow fully decentralization of the multisensory Kalman filter equations with state equality constraints among a number of sensing nodes to verify the data. The algorithm is developed from a centralized method named projection method to minimized the communication among nodes and can take place without any prior synchronization between nodes. Theoretical derivation is provided to the decentralized algorithm. Finally, the case study of the secondary chilled water pump system illustrates the effectiveness of the proposed method.

In this paper, DXF (Data Exchange File) file is used to extract the topological information of a distribution network. The basic information is configured to each CPN (Computing Process Node) and the local topology is obtained through broadcast communication. According to the topological sorting algorithm of the adjacency list, the global topology is obtained. Taking a typical building distribution network as an example, the feasibility of the method is analyzed and demonstrated.

With the wide application of BIM, the technology of Operation and Maintenance (O&M) management is developing. By combining BIM with facility management, it can visually manage the model and highly integrate information of the model. As a common data standard of BIM, IFC (Industry Foundation Classes) standard can avoid the problems caused by the mismatch of information between different models. This paper proposes a method of BIM-based facility management and describes the key technical issues in the development of O&M systems. The O&M system leverages BIMserver and its public plug-in interface as a system framework to analyze IFC data. Through the study in this paper, a new method can be developed for future underground space O&M management.

To solve the high labor and maintenance cost problems in actual engineering, a decentralized heating, ventilation, and air-conditioning (HVAC) system is configured following its physical layout. In a decentralized HVAC control system, each of the updated smart equipment can communicate with the adjacent nodes collaboratively to fulfill the load requirement. Furthermore, to achieve the global optimal operation of an HVAC system, a fully distributed constrained optimization is formulated. In this paper, a fully distributed genetic algorithm (GA) is developed to solve the proposed constrained optimization. The proposed method is confirmed to be effective to realize the global optimization of HVAC system through simulation study.

When an open-neutral fault occurs in an underground space low-voltage distribution system, the neutral-point voltage will offset, which will make a lot of equipment unable to work normally. In order to solve such problem, the changes of third harmonic current and neutral-point offset voltage were studied and analyzed, the parameters of Support Vector Machine (SVM) was optimized by using a genetic algorithm. The output of SVM whether normal or fault state can be distinguished by taking the variation of harmonic current and neutral offset voltage as the input of SVM. The experimental result shows that the method of optimizing the parameters of SVM based on the genetic algorithm can effectively detect the fault.