CN107680393A - Intelligent control method of crossroad traffic signal lamp based on time-varying domain - Google Patents

Abstract

The invention discloses an intelligent control method of a traffic signal light at a crossroad based on a time-varying universe. This method first uses the detector at the intersection to collect traffic data; based on the maximum queuing length in the current phase of the green light direction and the queuing length data in each phase of the red light direction, the congestion situation at the intersection is evaluated. In this case, the domain of discourse where the cycle length is obtained; under the domain of discourse of the cycle length, taking the maximum queue length on the current green light direction phase and the average parking times of each vehicle as input, the green light duration allocated by the current green light direction phase is Output, list the dynamic fuzzy rules; use the center of gravity method for clear calculation to obtain the green light duration; comprehensively consider the intersection traffic flow and driver safety, set limiting conditions, obtain the final green light duration, and complete the optimization of the timing scheme. The beneficial effect of the invention is that it can effectively enhance the traffic capacity of road intersections and reduce vehicle delay time.

Description

An intelligent control method for traffic lights at intersections based on time-varying domain of discourse

technical field

The invention belongs to the field of intelligent traffic systems, in particular to a method for real-time timing of traffic signal lights.

Background technique

The timing signal timing method mainly includes the British Webster (Webster) timing method in the world. On the basis of it, Australia's ARRB (Australian Road Research Board) timing method considers the supersaturated traffic situation and is a Webster delay model. amendments and extensions. The HCM (Highway Capacity Man) timing method in the United States is also widely used. In my country, there are conflict point method, critical lane method, estimation method, and Shanghai comprehensive algorithm. Among them, the Shanghai comprehensive algorithm is proposed on the basis of the existing signal timing methods at home and abroad, combined with the characteristics of my country's traffic, and its timing delay calculation is the same as the American HCM algorithm.

With the rapid development of the economy, the number of motor vehicles has increased sharply, and the traffic demand has expanded rapidly. The timing signal timing method cannot meet the requirements of efficiently improving urban traffic congestion.

With the expansion of the application scope and the increase of application areas of intelligent transportation systems, the real-time control and optimization of traffic signals has attracted the attention of more and more scholars at home and abroad. However, due to the strong randomness of the urban road traffic system, even in a relatively short period of time, the traffic parameters will change greatly, which brings great difficulties to the real-time timing of traffic lights. When describing the state of things, the time-varying universe and dynamic fuzzy rules solve the problem that it is difficult for people to define the membership function of fuzzy sets, and also provide a basis for the analysis of complex systems. method. Based on the time-varying universe, the present invention utilizes dynamic fuzzy rules to dynamically adjust the period length of traffic lights and the duration of green lights in real time, completes the real-time timing of traffic lights at intersections, effectively enhances traffic capacity at road intersections and reduces vehicle delay time.

Contents of the invention

The present invention provides a kind of intelligent control method of crossroad traffic lights based on time-varying universe, and the method comprises the following steps:

Step S1: Use the detector at the intersection to collect the required traffic data;

Step S2: Using the obtained data, based on the maximum queuing length in the current phase of the green light direction and the queuing length data of each phase in the current red light direction, evaluate the degree of congestion at the intersection, and obtain the cycle length under the current traffic flow conditions domain of discourse;

Step S3: Taking the current cycle length as the domain, taking the maximum queue length in the current green light direction phase and the average parking times of each vehicle as input, and the green light duration allocated by the current green light direction phase as output, list the dynamic fuzzy rules;

Step S4: Apply the center of gravity method to carry out clear calculation to obtain the duration of the green light;

Step S5: Judging the obtained green light duration, setting constraints, and completing the timing scheme.

Wherein said step S1 is further as follows: the required traffic data includes the maximum queue length gL on the phase of the current green light direction, the queue length rL _j on each phase of the current red light direction and the average parking times of each vehicle The average number of stops per vehicle is the average number of times that vehicles exiting the intersection encounter red lights in the current green light direction phase.

Wherein, said step S2 further comprises the following steps:

Step S21: The appropriate range of period length c is 40-180s, and the universe of period length c is set as Ω(t), and the universe of period length c under the time-varying universe is divided into continuous time-varying universe sequences: {Ω _k (t)}, (k∈N), Ω _k (t)=[0,40+20k], where k=1,2,...6,7;

Step S22: Use the 4-level service level suggested by Beijing Urban Planning and Design Institute, and use the average queuing length at red light as the basis to evaluate the degree of congestion at the intersection. According to its proposed 4-level service level, the queue length L at red lights is less than 50 meters for unobstructed VS, between 50 meters and 100 meters for relatively smooth NS, between 100 meters and 150 meters for relatively congested NJ, and more than 150 meters m is the congestion VJ; the digital form of the obtained gL and rL _j data is converted into the form of words VS, NS, NJ, VJ according to the proposed 4-level service level;

Step S23: According to the combination of base words, the domain of discourse of the cycle length under the current traffic flow is obtained.

Wherein, said step S3 further comprises the following steps:

Step S31: Input amount gL, The output volume T _l is fuzzy;

Step S32: The periodic length of the domain of discourse is a continuous time-varying domain of discourse sequence: {Ω _k (t)}, (k∈N), when the domain of discourse changes with time, the membership function of the fuzzy set defined on the domain of discourse It will also change with time, set input quantity gL and rL _j , output quantity on each Ω _k (t) respectively The membership function corresponding to each language value;

Step S33: Considering all possible situations, list the dynamic fuzzy rules R _k under the time-varying universe and when the universe of period length is Ω _k (t).

Wherein, said step S4 further includes the following steps:

Step S41: Obtain the fuzzy relationship matrix R _i from the dynamic fuzzy rules, and the total fuzzy rules are R;

Step S42: Obtain the fuzzy set corresponding to the output variable according to the general fuzzy relationship and the inference synthesis rule;

Step S43: Transform the fuzzy amount obtained by fuzzy reasoning into the clear amount in the current domain of discourse, perform clear calculation, and obtain the green light duration.

Wherein, said step S5 further comprises the following steps:

Step S51: Comprehensively consider the traffic flow at the intersection and the safety of drivers, so as to alleviate traffic congestion and reduce the phenomenon of red light running, and set a reminder. If the duration of the red light at a certain phase exceeds 130s, the system will receive a reminder 1;

Step S52: When the system receives the reminder 1, it adjusts the passing time of the current green light phase according to the forecasted traffic flow trend through short-term traffic flow prediction;

Step S53: T _l ≤ N _max , N _max is the maximum value of the passing time. According to the actual traffic flow setting at the intersection, the value should not exceed 80s. Set a reminder. If the green light duration exceeds N _max , the system will receive a reminder 2 ;

Step S54: When the system receives the reminder 2, it adjusts the passing time of the current green light phase and sets it as the maximum passing time.

The beneficial effects of the present invention are: by dynamically adjusting the cycle length of the traffic light and distributing the duration of the green light in real time, the number of vehicles passing through the intersection is increased, the maximum queuing length and the average queuing length are shortened, and the average delay of each period is reduced. The timing scheme effectively alleviates intersection congestion.

Description of drawings

Fig. 1 is a framework diagram of the real-time timing method for traffic lights at intersections of the present invention;

Fig. 2 is a flowchart of step S2 of the present invention;

Figure 3 is a schematic diagram of four-phase control at this intersection;

Fig. 4 is the membership function of gL when the period length discourse domain is Ω ₁ (t) and Ω ₂ (t);

Figure 5 shows the membership function of gL when the period length discourse domain is Ω ₃ (t) and Ω ₄ (t).

detailed description

The present invention will be described in detail below in conjunction with the accompanying drawings. It should be noted that the described embodiments are only intended to facilitate the understanding of the present invention, rather than limiting it in any way.

The invention provides a real-time timing method for traffic signal lights at intersections. As shown in Figure 1, specifically, the method includes the following steps:

Step S1: Use the detector at the intersection to collect the required traffic data;

The data that this method needs to collect includes the maximum queue length gL on the current phase of the green light direction, the queue length rL _j on each phase of the current red light direction, and the average number of parking times for each vehicle The average number of stops per vehicle is the average number of times that vehicles exiting the intersection encounter red lights in the current green light direction phase.

Step S2: Using the obtained data, based on the maximum queuing length in the current phase of the green light direction and the queuing length data of each phase in the current red light direction, evaluate the degree of congestion at the intersection, and obtain the cycle length under the current traffic flow conditions domain of discourse, the flow chart of the process is shown in Figure 2;

Step S21: The appropriate range of period length c is 40-180s. Let the universe of period length c be Ω(t), and the universe of period length c under the time-varying universe is divided into continuous time-varying universe sequences: {Ω _k ( t)},(k∈N), Ω _k (t)=[0,40+20k], where k=1,2,...6,7;

Step S22: Use the 4-level service level suggested by Beijing Urban Planning and Design Institute, and use the average queuing length at red light as the basis to evaluate the degree of congestion at the intersection. According to its proposed 4-level service level, the queue length L at red lights is less than 50 meters for unobstructed VS, between 50 meters and 100 meters for relatively smooth NS, between 100 meters and 150 meters for relatively congested NJ, and more than 150 meters m is the congestion VJ; the digital form of the obtained gL and rL _j data is converted into the form of words VS, NS, NJ, VJ according to the proposed 4-level service level;

Step S23: According to the combination of base words, the domain of discourse of the period length under the current traffic flow condition is obtained; taking the four-phase control at the intersection as an example, as shown in FIG. 3 . There are 35 cases of basic word combinations of the four phases. When the basic word combination forms are VS, VS, VS, VS//VS, VS, VS, NS, the domain of discourse of cycle length is Ω ₁ (t); the combination form is VS, VS, VS, NJ//VS, VS, VS, VJ//VS, VS, NS, NS//VS, VS, NS, NJ//VS, VS, NS, VJ//VS, VS, NJ , VJ, the universe of period length discourse is Ω ₂ (t); and so on, the corresponding length universe of each combination of base words is obtained.

Step S3: Under the domain of cycle length discourse, take the maximum queuing length on the current green light direction phase and the average parking times of each vehicle as input, and the green light duration allocated by the current green light direction phase as output, and list the dynamic fuzzy rules;

Step S31: Input amount gL, The fuzzification of the output volume T _l ; the variation range of gL is determined according to the actual traffic control experience. Assuming that the range of the input quantity is 0-200, the domain of discourse is {0,1,2,3,4,5,6,7,8,9,10}, and 7 fuzzy subsets are defined on the domain of discourse , the corresponding language values are {l ₁ (very short), l ₂ (short), l ₃ (short), l ₄ (medium), l ₅ (long), l ₆ (long), l ₇ (very long )}; set input The range of change is 0~4, the domain of discourse is {0,1,2,3,4,5}, five fuzzy subsets are defined on the domain of discourse, and the corresponding language value is {n ₁ (small), n ₂ (smaller), n ₃ (medium), n ₄ (larger), n ₅ (larger)}; suppose the range of output T _l is 0～40+20k, k=1,2,3,4 ,5,6,7, the domain of discourse is {0,1,2,3,4,5,6,7,8,9,10} and defines 5 fuzzy subsets on its domain of discourse, and the corresponding language value is {t ₁ (short), t ₂ (short), t ₃ (medium), t ₄ (long), t ₅ (long)};

Step S32: The periodic length of the domain of discourse is a continuous time-varying domain of discourse sequence: {Ω _k (t)}, (k∈N), when the domain of discourse changes with time, the membership function of the fuzzy set defined on the domain of discourse will also change with time, set the input quantity _gL , _The membership function corresponding to each language value of the output quantity T1; as shown in Figure 4, when the period length discourse domain is Ω ₁ (t) and Ω ₂ (t), the membership function corresponding to each language value of the input quantity gL; As shown in Figure 5, when the cycle length universe is Ω ₃ (t) and Ω ₄ (t), the membership function corresponding to each language value of the input quantity gL; similarly when the cycle length universe is Ω ₅ (t), When Ω ₆ (t) and Ω ₇ (t), the membership function corresponding to each language value of the input amount gL is similar to Fig. 4 and Fig. 5 respectively; similarly, defining the input amount The membership function corresponding to each language value of the output quantity T _l on each Ω _k (t);

Step S33: Considering all possible situations, list the dynamic fuzzy rules R _k under the time-varying universe and when the universe of period length is Ω _k (t).

Step S4: Apply the center of gravity method to carry out clear calculation to obtain the duration of the green light;

Step S41: Obtain the fuzzy relationship matrix R _i from the dynamic fuzzy rules, and the total fuzzy rules are R; in the grammar of fuzzy language, the rules are generally expressed in the sentence pattern of if (rule antecedent) then (conclusion). Each language control rule corresponds to a fuzzy relation The fuzzy relationship matrix R _i can be obtained from the fuzzy control rules, and the total fuzzy relationship is

Step S42: Obtain the fuzzy set of the corresponding output variable according to the general fuzzy relationship and inference synthesis rules:

Step S43: Transform the fuzzy amount obtained by fuzzy reasoning into the clear amount in the current domain of discourse, perform clear calculation, and obtain the green light duration.

Step S5: Judging the obtained green light duration, setting constraints, and completing the timing scheme.

Step S51: Comprehensively consider the traffic flow at the intersection and the safety of drivers, so as to alleviate traffic congestion and reduce the phenomenon of red light running, and set a reminder. If the duration of the red light at a certain phase exceeds 130s, the system will receive a reminder 1;

Step S52: When the system receives the reminder 1, it adjusts the passing time of the current green light phase according to the forecasted traffic flow trend through short-term traffic flow prediction;

Step S53: T _l ≤ N _max , N _max is the maximum value of the passing time. According to the actual traffic flow setting at the intersection, the value should not exceed 80s. Set a reminder. If the green light duration exceeds N _max , the system will receive a reminder 2 ;

Step S54: When the system receives the reminder 2, it adjusts the passing time of the current green light phase and sets it as the maximum passing time.

The above is only a specific implementation mode in the present invention, but the scope of protection of the present invention is not limited thereto. Anyone familiar with the technology can understand the conceivable transformation or replacement within the technical scope disclosed in the present invention. All should be covered within the scope of the present invention. Therefore, the protection scope of the present invention should be determined by the protection scope of the claims.

Claims (6)

1. it is a kind of based on when variable universe crossroad access signal lamp intelligent control method, it is characterised in that this method includes Following steps：

Step S1：Using the detector of crossroad, traffic data needed for collection；

Step S2：With the data obtained, with the maximum queue length in the phase of current green light direction and the current each phase in red light direction Queue length data on position are foundation, to evaluate intersection congestion level, are drawn in the case of current flows, Cycle Length Domain；

Step S3：Under the Cycle Length domain, with the maximum queue length in the phase of current green light direction and each car Average stop frequency is input, and the long green light time of current green light direction phase assignments is output, lists dynamic fuzzy rule；

Step S4：Sharpening calculating is carried out using gravity model appoach, draws long green light time；

Step S5：Gained long green light time is judged, constraints is set, completes timing scheme.

2. according to the method for claim 1, it is characterised in that the step S1 is further：Required traffic data bag Include the maximum queue length gL in the phase of current green light direction, the queue length rL in the current each phase in red light direction_jIt is and every The average stop frequency of car

3. according to the method for claim 1, it is characterised in that the step S2 further comprises the steps：

Step S21：Cycle Length c 40~180s of OK range, if the domain of Cycle Length is Ω (t), by when variable universe Lower Cycle Length c domain is divided into consecutive hours variable universe sequence：{Ω_k(t)},(k∈N)；

Step S22：The 4 grades of service levels suggested using urban planning and design academy of Beijing, to be averagely lined up during red light therein Length item is foundation, to evaluate intersection congestion level；By obtained gL and rL_jThe digital form of data, according to the 4 of suggestion grades Service level, be converted to word VS, NS, NJ, VJ form；

Step S23：According to the combined situation of keyword, draw in the case of current flows, the domain of Cycle Length.

4. according to the method for claim 1, it is characterised in that the step S3 further comprises the steps：

Step S31：Input quantity gL,Output quantity T_lBlurring；

Step S32：Cycle Length domain is consecutive hours variable universe sequence：{Ω_k(t) }, (k ∈ N), when domain becomes with the time During change, the membership function for the fuzzy set being defined on domain can also change with the time, respectively in each Ω_k(t) set on Put input quantity gL and rL_j, output quantityMembership function corresponding to each Linguistic Value；

Step S33：Consider all possible situation, when listing under variable universe, the domain of Cycle Length is Ω_k(t) dynamic analog when Paste regular R_k。

5. according to the method for claim 1, it is characterised in that the step S4 further comprises the steps：

Step S41：Fuzzy relationship matrix r is obtained by dynamic fuzzy rule_i, total fuzzy rule is R；

Step S42：The fuzzy set of corresponding output variable is obtained according to total fuzzy relation and push-pull picklingline；

Step S43：The fuzzy quantity that fuzzy reasoning is obtained is converted into the clear amount in present domain, carries out sharpening calculating, obtains Go out long green light time.

6. according to the method for claim 1, it is characterised in that the step S5 further comprises the steps：

Step S51：Consider crossroad access flow and the current safety of driver, alleviate traffic congestion and reduce and make a dash across the red light Phenomenon, set and remind, if certain phase red light duration, more than 130s, system receives prompting 1；

Step S52：When system receives prompting 1, by the prediction of short-term traffic flow, according to the variation tendency of the traffic flow of prediction, Adjust the transit time of current green light phase；

Step S53：T_l≤N_max, N_maxFor the maximum of transit time, set according to the magnitude of traffic flow at actual crossing, the value is best 80s is not exceeded, sets and reminds, if long green light time is more than N_max, then system receive prompting 2；

Step S54：When system receives prompting 2, current green light phase transit time is adjusted, it is transit time maximum to set it.