Modeling Multi-commodity Trade: Information Exchange Methods

In the chapter, first we discuss needs and motivations for market-based mechanism for complex, distributed systems, especially focusing on infrastructures sectors. A great number of issues related to market mechanisms developing is modeled by decisions models. There are many completely different attempts to organizing the mechanisms of finding market equilibrium. These attempts are hard to reliably compere on the common test cases due to huge effort needed for numerical data adaptation to each approach. We give strong foundations for Multicommodity Market Model (M³), which describes the data on the input and output sides of the market clearing process. Its main advantage is very high expressiveness. M³ covers most known trade mechanisms, including different auction types, multi-commodity mechanisms, bilateral contracts and others. In the chapter we present architecture of M³ , including formal model.

A detailed data model for multi-commodity trade is presented, based on formal model described in previous chapter. This data model has form of UML class diagrams, and widely uses generic (i.e. metadata driven) constructs. On the basis of the UML model, an XML dialect, called M3-XML, is proposed, which can be used to store formal descriptions of multi-commodity market and offers, or as a content language for market data interchange.

Multi-commodity trading can be performed both in centralized and distributed architecture environment. Choice of architecture creates a need to choose an appropriate communication model. Thus the definition of communication models is required. Foundation for Intelligent Physical Agents (abbrev. FIPA) organization, supplies us with communication models, appropriate for multi-agent systems. In this chapter we describe these models, and we also propose how FIPA standards can be enriched to multi-commodity and distributed elements.

This chapter analyzes possibilities of integration between Web services and multi-agent technology. Efforts of Agents and Web Services Interoperability Working Group (AWSI WG) are described, which is focused on Web Services and FIPA (Foundation for Intelligent Physical Agents) interoperability. A hybrid architecture for conducting trade in a multi-commodity markets, which is based on multi-agent approach combined with the best practices and standards of the Service Oriented Architecture is also proposed. The architecture allows the trade to be conducted by large parties with well structured and defined offer as well as smaller entities, which operate on smaller scale and do not have resources to build full featured catalogues.

In this chapter we concern the trading platform as the system of independent software agents, which communicate with each other. Agents’ interaction lead them to the exchange of commodities. Applying proper communication language ensures compliance with relevant trading mechanism. Such an approach leads us to the design and implementation of multi-agent platforms for multi-commodity trade. Implementations in Java language and in AIMMS environment are described.

In the constantly expanding Semantic Web, the domain of describing offers and concepts related to them plays an important role. The use of ontologies provides a means for automated processing of knowledge, which is expressed in a standardized representation language. We argue that the M³ model can be represented in OWL, a standard ontology language. We show the benefits of such a representation. Several ontologies are shown, which can complement the descriptions made in M³ . We propose a translation of the M³ model and the M3-XML data format into an OWL-based ontology.We also show exemplary reasoning tasks, which could be accomplished with standard description logic reasoners. Some problems with using ontologies for expressing offers are also described.

In this paper several reliability aspects of multi-commodity trade are discussed. As it is very often required that trading decisions should be taken both rationally and very quickly, e.g on a short-term electric power market, in such cases trading is performed automatically by multi-agent systems. Thus the question of MAS reliability arises, which is herein considered in two aspects – topological and functional. Interactions of agents in a MAS occur according to a certain topological pattern which can be directly transformed to the structural reliability model of that MAS. A number of such topological patterns are presented along with respective reliability models. In its functional aspect a MAS can be seen as a graph whose nodes process input information and pass it to other nodes thus fulfilling collectively certain task. An agent’s malfunction can lead to a delay, misfulfillment or failure of that task. An interesting model of inter-agent functional dependence based on game-theoretical approach is presented. Apart from reliable execution of trading operations, an important issue is the ability to quickly assess whether the fulfillment of a contract is technically possible. An example is given where the effective conditions which energy suppliers’ capabilities must meet are defined as the constraints of a transportation problem. Also, a non-trivial problem of quickly finding those constraints is addressed.

The electrical energy markets illustrate very well complexity of the infrastructure markets. Achieving the market balance, while maintaining its high efficiency, is a challenging task which requires both a rich commodities structure and a complex structure of the market processes. Market balancing is being obtained by multi-step processes of the multi-commodity trade. In this chapter we show that the fulfillment of various specific electricity market requirements is possible by applying model M³ . We focus on the balancing segment of the market and present relations and mappings between various concepts of a real balancing market and notions of M³ . We discuss some potential ways of market development to demonstrate how M³ notions facilitate modeling of the new solutions.

Uncertainties of pollution inventories are often high due to low precision of emission quantity assessments for many emitting sources. A good example is emission of greenhouse gases, where uncertainty of some sources may be as high as 40-100%, while uncertainty of other sources is as low as 2-3%. This discrepancy in uncertainty should be accounted for in compliance as well as in emissions trading, because the traded commodities have different quality. The compliance and emissions trading rules have been discussed in earlier papers by the present authors [14, 15, 16, 17]. In this chapter we focus on presentation of the idea of a market for emissions with so highly scattered uncertainties.

Greenhouse gases emission permits trading can be modeled using the multi-agent platform for multi-commodity exchange. A simulation of this kind of trade is described in the paper. A party can use one of two strategies to find a good partner to achieve best gain: (i) bilateral trade with a randomly chosen feasible partner, (ii) a tender. In the tender trade, parties submit offers to the current tender operator; the tender operator chooses the offer of the party that maximizes his gain. The results of simulation are presented.

In this chapter the Multi-commodity Market Model is applied for modelling virtual network market data. Two different auction models supporting VPN service trading are considered. The first one enables to specify VPN demand requirements by the means of pipe model, while the second one leverages the hose model to describe VPN service in the buy offer. The market data for both auctions can be easily implemented in M³ using the bundled offer and network virtualization concepts.