1 Introduction
Authors | Opt. technique | Contribution |
---|---|---|
Lin et al. (2022) | Fuzzy logic and differential evolution algorithm (FASM-MDEA) | Optimization of phase plans and signal timing for six intersections |
Zhang et al. (2022) | Multi-objective non-dominated sorting genetic algorithm (NSGA-III) | Signal cycle and green time optimization |
Islam et al. (2022) | Convolutional neural network and long short-term memory (CNN-LSTM) | Signal phasing and timing optimization for seventeen intersections |
Liu et al. (2022) | Mixed integer linear programming (MILP) | Signal timing optimization |
Zhao et al. (2022) | Reinforcement learning (RL) | Optimization of phase duration of traffic signals for sixteen intersections |
2 Materıal and methods
2.1 Flower pollination algorithm
2.2 Crow search algorithm
-
Crows live in the form of flock.
-
Crows memorize the position of their hiding places.
-
Crows follow each other to do thievery.
-
Crows protect their caches from being pilfered by a probability.
2.3 Type-2 fuzzy logic
2.4 Traffic simulation program
2.4.1 Vehicle creating module
Vehicle types | Lengths (m) |
---|---|
Car | 5,00 |
Minibus | 6,00 |
Truck | 8,50 |
Bus | 12,00 |
Lorry | 13,50 |
2.4.2 Vehicle dynamics module
Vehicle types | Speed Ranges | |||||||||
---|---|---|---|---|---|---|---|---|---|---|
0–10 | 10–20 | 20–30 | 30–40 | 40–50 | 50–60 | 60–70 | 70–80 | 80–90 | 90–100 | |
Car | 2,44 | 2,74 | 1,83 | 1,52 | 1,52 | 1,52 | 1,22 | 0,91 | 0,61 | 0,61 |
Minibus | 1,43 | 1,63 | 1,51 | 1,06 | 0,94 | 0,80 | 0,65 | 0,52 | 0,39 | 0,26 |
Truck | 0,86 | 0,74 | 0,66 | 0,62 | 0,53 | 0,43 | 0,34 | 0,25 | 0,17 | 0,09 |
Bus | 2,28 | 1,62 | 0,97 | 0,81 | 0,70 | 0,50 | 0,43 | 0,29 | 0,23 | 0,15 |
Lorry | 0,75 | 0,65 | 0,43 | 0,34 | 0,26 | 0,19 | 0,13 | 0,09 | 0,04 | 0,00 |
2.4.3 Signal control module
2.4.4 Graphical user interface
2.5 Hybrid traffic signal controller
2.5.1 Phase plan optimization module
Objective function | \(f={\text{min}}\left\{\frac{\sum_{l=1}^{k}{d}_{l}}{\sum_{l=1}^{k}{q}_{l}}\right\}\) Figure 5: Flow chart of phase optimization module | |
Decision variable | gi | g is green light duration i is number of phase |
Constraints | 8 \(\le \) gi \(\le \) 60 | |
\(0\le \frac{{q}_{i}\times C}{s\times {g}_{i}}\le 1.4\) |
2.5.2 Signal timing optimization module
Objective function | \(f={\text{min}}\left\{\sum_{l=1}^{k}{d}_{n}\right\}\) | |
Decision variables | gchanges | gchanges is green time change amount |
MFUi | MFUi is FL upper membership function center | |
bi | bi is FL upper membership function left leg | |
bj | bj is FL upper membership function right leg | |
MFLi | MFLi is FL lower membership function center | |
bii | bii is FL lower membership function left leg | |
bjj | bjj is FL lower membership function right leg | |
Constraints | − 6 \(\le \) gchanges \(\le \) +6 | |
MFUi \(\le \) MFUi+1 | ||
MFUi \(=\) MFLi | ||
bi \(\le \) MFUi | ||
bj \(\ge \) MFUi | ||
bi \(\le \) bii \(\le \) MFUi | ||
MFUi \(\le \) bjj \(\le \) bj |
-
h0: No difference between averages.
-
h1: At least one average different from the others.
Traffic Situations | T | Traffic volumes | − 4 + 4 | − 6 + 6 | − 8 + 8 | − 10 + 10 | ||
---|---|---|---|---|---|---|---|---|
V1 | V2 | V3 | Delay | Delay | Delay | Delay | ||
Low | 15 | 400 | 400 | 400 | 34.01 | 32.51 | 37.88 | 38.70 |
Medium | 15 | 800 | 800 | 800 | 86.00 | 78.65 | 83.70 | 89.77 |
High | 15 | 1200 | 1200 | 1200 | 97.82 | 95.60 | 98.18 | 101.09 |
Traffic Situations | T | Traffic volumes | − 4 + 4 | − 6 + 6 | − 8 + 8 | − 10 + 10 | |||
---|---|---|---|---|---|---|---|---|---|
V1 | V2 | V3 | V4 | Delay | Delay | Delay | Delay | ||
Low | 15 | 400 | 400 | 400 | 400 | 71.33 | 66.53 | 76.13 | 79.59 |
Medium | 15 | 800 | 800 | 800 | 800 | 86.81 | 84.88 | 86.35 | 88.50 |
High | 15 | 1200 | 1200 | 1200 | 1200 | 108.68 | 103.13 | 105.44 | 109.84 |
Samples | Average | Standard deviation | Levene homogeneity test | ANOVO | ||
---|---|---|---|---|---|---|
F | P | |||||
− 4 + 4 | 6 | 80.77 | 26.11 | 0.992 | 0.096 | 0.962 |
− 6 + 6 | 6 | 76.88 | 25.22 | |||
− 8 + 8 | 6 | 81.28 | 23.71 | |||
− 10 + 10 | 6 | 84.58 | 24.83 | |||
Total | 24 | 80.88 | 23.46 |
Membership function | Verbal expression | Value |
---|---|---|
MF1 | Extreme low (EL) | − 6 |
MF2 | Very low (VL) | − 4 |
MF3 | Low (L) | − 2 |
MF4 | Normal (N) | 0 |
MF5 | High (H) | 2 |
MF6 | Very high (VH) | 4 |
MF7 | Extreme high (EH) | 6 |
2.5.3 Phase sequence optimization module
3 Intersections parameters for simulations and results
-
Each leg has two lanes, with 3.6 m lane width and 0 slope.
-
Start-up lost time is 3.6 s.
-
Free flow speed is 50 km/h.
-
Lost time for yellow light is 2 s and for all-red is 1 s.
-
Right turns and left turns were determined to be 10% and 20%, respectively.
-
Saturation flow was set at 1800 veh/h/lane.
State | Approach pattern | West (veh/h) | East (veh/h) | North (veh/h) |
---|---|---|---|---|
Low | All-equal | 100 | 100 | 100 |
All-equal | 400 | 400 | 400 | |
Mixed | 100 | 250 | 400 | |
Medium | All-equal | 500 | 500 | 500 |
All-equal | 800 | 800 | 800 | |
Mixed | 650 | 800 | 800 | |
High | All-equal | 900 | 900 | 900 |
All-equal | 1200 | 1200 | 1200 | |
Mixed | 900 | 1050 | 1050 |
State | Approach pattern | West (veh/h) | East (veh/h) | North (veh/h) | South (veh/h) |
---|---|---|---|---|---|
Low | All-equal | 100 | 100 | 100 | 100 |
All-equal | 400 | 400 | 400 | 400 | |
Mixed | 100 | 250 | 400 | 400 | |
Medium | All-equal | 500 | 500 | 500 | 500 |
All-equal | 800 | 800 | 800 | 800 | |
Mixed | 650 | 800 | 800 | 500 | |
High | All-equal | 900 | 900 | 900 | 900 |
All-equal | 1200 | 1200 | 1200 | 1200 | |
Mixed | 900 | 1050 | 1050 | 1200 |
Algorithm | Three-leg | ||||
---|---|---|---|---|---|
N | Mean | Std. deviation | Minimum | Maximum | |
Fixed-time | 9 | 47.3293 | 22.61072 | 10.95 | 77.67 |
Type-1FL-TSC | 9 | 43.0452 | 20.66746 | 11.30 | 68.74 |
FPA-TSC | 9 | 43.0172 | 20.94332 | 8.05 | 67.23 |
HTSC | 9 | 42.4345 | 20.59502 | 8.50 | 65.85 |
Algorithm | Three-leg with left turn pocket | ||||
---|---|---|---|---|---|
N | Mean | Std. deviation | Minimum | Maximum | |
Fixed-time | 9 | 62.9787 | 18.29680 | 31.56 | 85.87 |
Type-1FL-TSC | 9 | 47.5462 | 18.88301 | 16.00 | 70.58 |
FPA-TSC | 9 | 47.0178 | 18.36277 | 16.02 | 69.34 |
HTSC | 9 | 44.5993 | 16.97585 | 14.27 | 64.61 |
Algorithm | Four-leg | ||||
---|---|---|---|---|---|
N | Mean | Std. deviation | Minimum | Maximum | |
Fixed-time | 9 | 61.8995 | 18.66233 | 31.09 | 81.53 |
Type-1FL-TSC | 9 | 50.9896 | 19.99126 | 18.06 | 79.15 |
FPA-TSC | 9 | 50.5641 | 19.48156 | 18.66 | 77.75 |
HTSC | 9 | 48.5330 | 19.57509 | 16.93 | 75.43 |
Algorithm | Four-leg with left turn pocket | ||||
---|---|---|---|---|---|
N | Mean | Std. deviation | Minimum | Maximum | |
Fixed-time | 9 | 69.8196 | 13.21182 | 44.30 | 83.19 |
Type-1FL-TSC | 9 | 53.3215 | 16.47929 | 26.82 | 75.53 |
FPA-TSC | 9 | 52.6630 | 15.97809 | 27.79 | 74.71 |
HTSC | 9 | 51.8187 | 15.40894 | 27.08 | 72.34 |
Algorithm | Three-leg | Three-leg with left turn pocket | Four-leg | Four-leg with left turn pocket | ||||
---|---|---|---|---|---|---|---|---|
Mean rank | p values | Mean rank | p values | Mean rank | p values | Mean rank | p values | |
Fixed-time | 3.89 | 0.001 | 4.00 | 0.000 | 4.00 | 0.000 | 4.00 | 0.000 |
Type-1FL-TSC | 2.44 | 2.67 | 2.78 | 2.78 | ||||
FPA-TSC | 2.22 | 2.22 | 2.22 | 2.00 | ||||
HTSC | 1.44 | 1.11 | 1.00 | 1.22 |
Intersections | MPE | |||||
---|---|---|---|---|---|---|
Fixed-time | Fixed-time | Fixed-time | Type-1FL-TSC | Type-1FL-TSC | FPA-TSC | |
/ | / | / | / | / | / | |
Type-1FL-TSC | FPA-TSC | HTSC | FPA-TSC | HTSC | HTSC | |
Three-leg | − 8.64 | − 10.73 | − 11.59 | − 1.67 | − 2.70 | − 0.80 |
Three-leg with left turn pocket | − 27.23 | − 27.86 | − 31.45 | − 0.71 | − 5.70 | − 5.04 |
Four-leg | − 19.81 | − 20.27 | − 23.99 | − 0.42 | − 5.26 | − 4.83 |
Four-leg with left turn pocket | − 25.17 | − 25.97 | − 27.11 | − 0.95 | − 2.42 | − 1.47 |