A highway patrol path optimization method and system in an uncertain environment and a storage medium

Abstract

The application provides a highway patrol path optimization method, system and storage medium in an uncertain environment, the method is: taking the real-time average driving speed of a patrol vehicle and the estimated processing time of a sudden event as uncertain parameters to construct a patrol scene set; obtaining the minimum total cost and the maximum total priority value of events to be responded to in each scene in the patrol scene set through a global dynamic path optimization model; constructing a patrol path robust optimization model by taking the minimum total cost expectation value in all scenes and the maximum total priority expectation value of events to be responded to in all scenes; solving the patrol path robust optimization model based on the sum in each scene, outputting an optimal robust solution set, and selecting an optimal patrol path from the optimal robust solution set. The method improves the feasibility of the patrol path, and takes into account the requirements of sudden event processing and completion of all road section inspections.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of path optimization, in particular to a highway patrol path optimization method, system and storage medium under uncertain environment. BACKGROUND

[0002] The generation of the dynamic path of the highway patrol has a certain advance in time, that is, the patrol vehicle updates the path after receiving the sudden event in the patrol process, but the vehicle may still encounter various random events such as traffic accidents, obstacle cleaning, temporary construction work and the like along the way during the driving process according to the dynamic path, causing traffic congestion. Influenced by random events, the patrol vehicle is difficult to continuously patrol according to the specified patrol driving speed, affecting the efficiency of the patrol work, and even failing to complete the specified patrol work plan. In addition, although the patrol workers can respond to the sudden events, due to the complexity of the events and the difference in individual processing capacity, there will still be a certain fluctuation in the actual processing time, so it is difficult to completely and accurately describe the real processing situation by relying on manual identification of the event type and presetting the processing time, and other influencing factors need to be further considered.

[0003] For the path responding to real-time sudden events in the patrol process, the feasibility of the path relates to whether the sudden event responded can be processed in time according to the established path, thereby affecting the traffic safety of the road. The Chinese patent with publication number CN120146357B discloses a highway patrol path optimization method, device and storage medium, which generates an initial patrol path by constructing a highway basic data set, receives sudden event data in real time during the patrol process, and divides the response mode according to the position, and then divides the priority of the sudden event to be responded and generates a dynamic patrol path. Although this method can balance the balance between the operation time and the priority of the sudden event in the path, it does not consider the influence of uncertain factors in the driving process, which may lead to the fact that the path scheme cannot achieve the established target.

[0004] In summary, there is an urgent need for a highway patrol path optimization method, system and storage medium under uncertain environment to solve the problems existing in the prior art. SUMMARY

[0005] The present application aims to provide a highway patrol path optimization method under uncertain environment, which aims to solve the problem that the prior art does not consider the influence of the fluctuation of the processing time of various sudden events in the patrol process and the fluctuation of the average driving speed of the patrol vehicle in the driving process, which leads to the fact that it is difficult to completely and accurately describe the real processing situation, and there is a problem that the specified patrol work plan cannot be completed. The specific technical scheme is as follows:

[0006] A highway patrol path optimization method under uncertain environment, comprising:

[0007] acquire basic data of highway patrol in jurisdiction, real-time traffic running data and real-time patrol data of highway in jurisdiction;

[0008] take real-time average driving speed of patrol vehicle and estimated processing time of sudden event as uncertain parameters, construct patrol scenario set ;

[0009] take total cost of patrol path as first objective function, maximize total priority value of events to be responded in patrol path as second objective function, construct global dynamic path optimization model; obtain minimum total cost and maximum total priority value of events to be responded in each scenario in patrol scenario set through global dynamic path optimization model , represent the i-th scenario in patrol scenario set ;

[0010] take minimum total cost expectation value in all scenarios and maximum total priority expectation value of events to be responded in all scenarios, construct patrol path robust optimization model; solve patrol path robust optimization model based on and , output optimal robust solution set;

[0011] select optimal patrol path from optimal robust solution set and send to patrol vehicle.

[0012] Preferably, take real-time average driving speed of patrol vehicle and estimated processing time of sudden event as uncertain parameters to construct patrol scenario set , specifically:

[0013] select regulated driving speed of patrol vehicle and benchmark estimated processing time of single sudden event as benchmark scenario; set

[0014] fluctuation driving speeds less than regulated driving speed , take regulated driving speed and fluctuation driving speeds as real-time average driving speed to construct speed scenario set ; wherein, the set of fluctuation driving speeds is represented as , is adjustment coefficient, ​​​​​​is an integer, is an integer greater than or equal to 1;

[0015] setting a fluctuation processing time , the reference estimated processing time and a fluctuation processing time are all used to construct a time scenario set ; wherein the set of fluctuation processing times is expressed as , there are such that and ; , , all represent adjustment coefficients, all are integers, is an integer greater than or equal to 2;

[0016] The speed scenario set and the time scenario set are combined to generate a patrol scenario set , which is further converted into the set , the set and the set , wherein the patrol scenarios contained in the sets , , represent the patrol scenario composed of the A th real-time average travel speed in the set and the B th estimated processing time in the set .

[0017] Preferably, the patrol path robust optimization model is expressed as:

[0018]

[0019]

[0020] In formulas (1.7) and (1.8), is a patrol scenario set, is any scenario in the patrol scenario set, is the occurrence probability of the scenario , and is the minimum total cost of the feasible solution under the scenario . is the maximum total priority value of the events to be responded under the scenario .

[0021] Preferably, the scenario probability constraints of the robust optimization model for the patrol path are expressed as follows:

[0022]

[0023] The robust constraints of the robust optimization model for the patrol path are expressed as follows:

[0024]

[0025]

[0026] In formulas (1.10) and (1.11), The regret factor is... The range of values ​​is .

[0027] Preferred, based on each scenario and Solve the robust optimization model for the patrol path and output the optimal robust solution set, specifically:

[0028] Calculate the minimum total cost under each scenario. set and the total priority value of the maximum pending events set ;in, Let A represent the minimum total cost under the patrol scenario consisting of the A-th real-time average driving speed and the B-th estimated processing time. This represents the total priority value of the maximum pending events in the patrol scenario, which consists of the Ath real-time average driving speed and the Bth estimated processing time.

[0029] The NSGA-II algorithm is used to solve the robust optimization model of the inspection path, and the optimal robust solution set is output. And the set of expected total costs corresponding to the optimal robust solution set. and the set of expected total priority values ​​for events to be responded to ;in, Represents the optimal robust solution set The first in A robust solution, express The expected total cost express The expected total priority of the events to be responded to.

[0030] Preferably, the NSGA-II algorithm solves the robust optimization model for the inspection path as follows:

[0031] A1. Setting Population Size Number of chromosome genes Maximum number of iterations Crossover probability and mutation probability ;

[0032] A2. Generate the initial population ;

[0033] A3. Calculate each individual In patrol scenario set Total cost and each individual In patrol scenario set Total priority value of pending events ;in, , individual The minimum total cost under the patrol scenario consisting of the A-th real-time average driving speed and the B-th estimated processing time. Represents an individual The maximum total priority value of events awaiting response in the patrol scenario consisting of the Ath real-time average driving speed and the Bth estimated processing time;

[0034] A4. For each individual Determine whether the regret constraint is satisfied: If and Then it is considered that the individual The regret constraint is satisfied at this time. , ,otherwise ;

[0035] in, Represents an individual The expected total cost Represents an individual The expected total priority of the events to be responded to. For the context Probability of occurrence, scenario Refers to the first The real-time average driving speed and the first The inspection scenario consists of a number of estimated processing times. Represents an individual In the context The minimum total cost, Represents an individual In the context The maximum total priority value of pending events;

[0036] A5. Let the fitness value of the expected total cost be _____. The fitness value of the expected total priority of the events to be responded to is ;

[0037] A6. Determine if the maximum number of generations has been reached. If the fitness value is reached, proceed to step A7; otherwise, adjust the population based on the fitness value of each individual. Perform non-dominated sorting and crowding calculation, execute selection, crossover, and mutation operations, merge the populations, and then perform non-dominated sorting and crowding calculation again to generate a new population. And re-enter step A3; in non-dominated sorting, if there exists or Then let the solution be found. For the subordinating solution;

[0038] A7. Using the current population as the optimal robust solution set, output the optimal robust solution set, the expected total cost set corresponding to the optimal robust solution set, and the expected total priority set of the events to be responded to.

[0039] Preferably, when the number of solutions in the optimal robust solution set is two or more, the selection of the optimal inspection path is specifically as follows:

[0040] The expected total cost of each solution and the expected total priority of the events to be responded to in the optimal robust solution set are normalized.

[0041] For each solution, the normalized expected total cost and the expected total priority of the events to be responded to are weighted and calculated, and then selected. The solution with the smallest value is taken as the optimal inspection path;

[0042]

[0043] in: To solve The normalized value of the corresponding expected total cost. To solve The normalized value of the expected total priority of the corresponding events to be responded to. The weighting coefficients for the expected total cost. The weighting coefficients represent the expected total priority of the events to be responded to. ;

[0044] If there is only one solution in the optimal robust solution set, then that solution is selected as the optimal inspection path.

[0045] This invention also provides a highway patrol route optimization system under uncertain environments. This system employs the aforementioned patrol route optimization method and includes:

[0046] The data acquisition and storage module is used to acquire and store basic data of highway patrol within the jurisdiction, real-time traffic operation data of highways within the jurisdiction, real-time inspection data, and regret coefficient.

[0047] The patrol scene generation module is used to construct a patrol scene set by taking the real-time average driving speed of the patrol vehicle and the estimated processing time of the sudden event as uncertain parameters ;

[0048] The global dynamic path optimization model solution module is used to obtain the minimum total cost under each scene And the maximum total priority value of the events to be responded ;

[0049] The patrol path robust optimization model solution module is used to obtain the minimum total cost and the maximum total priority value of the events to be responded corresponding to each feasible solution under each scene by calling the global dynamic path optimization model solution module, and output an optimal robust solution set with the minimum total cost under all scenes and the maximum total priority value of the events to be responded under all scenes as the target, and filter out the optimal patrol path

[0050] The information feedback module is used to feed back the optimal patrol path to the remote control end and display it on the screen of the patrol vehicle.

[0051] The application further provides a highway patrol path optimization system under an uncertain environment, comprising a memory and a processor, the memory stores a computer program, and the processor executes the computer program to perform the patrol path optimization method.

[0052] The application further provides a storage medium, which stores a computer program, and the computer program is executed to perform the patrol path optimization method.

[0053] The technical scheme of the application has the following beneficial effects:

[0054] The patrol path optimization method uses robust optimization theory, considers that the actual average driving speed of the staff during the patrol process will be lower than the specified vehicle speed due to random events along the way, and that the actual processing time of a single sudden event has uncertainty, takes the real-time average driving speed of the patrol vehicle and the estimated processing time of a single sudden event as uncertain parameters, clearly distinguishes the fluctuation scenes of the real-time average driving speed of the patrol vehicle and the fluctuation scenes of the estimated processing time of a single sudden event, so as to ensure that various scenes that may occur during the patrol process can be covered as much as possible. Meanwhile, a patrol path robust optimization model is further constructed on the basis of the global dynamic path optimization model, with the minimum total cost under all scenes and the maximum total priority value of the events to be responded under all scenes as the target, and the optimal patrol path under the current real-time patrol data (i.e. the current road section to be patrolled and the sudden event) is solved through the patrol path robust optimization model, so as to improve the feasibility of the patrol path, and meet the requirements of sudden event processing and completing all road section patrols.

[0055] In addition to the above described objects, features and advantages, the present application has other objects, features and advantages. These will become apparent from the following detailed description of the application, taken in conjunction with the accompanying drawings. BRIEF DESCRIPTION OF DRAWINGS

[0056] The accompanying drawings are included to provide a further understanding of the application, and are incorporated in and constitute a part of this specification. The drawings illustrate exemplary embodiments of the application and, together with the description, serve to explain the application without imposing undue limitation thereof. In the drawings:

[0057] Figure 1 is a flow chart of the method for optimizing the patrol path of the expressway in an uncertain environment in Example 1;

[0058] Figure 2 is a directed graph of the expressway network in Example 2;

[0059] Figure 3 is a schematic diagram of the positions of the inspection vehicle and the road sections to be inspected in the directed graph in Example 2;

[0060] Figure 4 is a schematic diagram of the positions of the inspection vehicle, the road sections to be inspected and the emergency event in the directed graph in Example 2;

[0061] Figure 5 is a schematic diagram of the patrol path generated by using the method of Example 1 in Example 2;

[0062] Figure 6 is a schematic diagram of the patrol path generated by using the prior art in Example 2. DETAILED DESCRIPTION

[0063] In order to facilitate the understanding of the present application, the present application will be described more fully below, and preferred embodiments of the present application will be given. However, the present application can be realized in many different forms and is not limited to the embodiments described herein. On the contrary, these embodiments are provided so that the disclosure of the present application can be more thorough and comprehensive.

[0064] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the present application belongs. The terminology used in the description of the application herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the application.

[0065] Example 1:

[0066] A highway patrol path optimization method, device and storage medium are disclosed in Chinese Patent Publication No. CN120146357B. The planning of the patrol path is completed by judging the type of the sudden event and presetting its processing time. The influence of the fluctuation of the processing time of various sudden events in the patrol process and the fluctuation of the average driving speed in the driving process of the inspection vehicle is not considered, which makes it difficult to completely and accurately describe the real processing situation and causes the problem of being unable to complete the prescribed patrol work plan.

[0067] To this end, the present embodiment provides a highway patrol path optimization method in an uncertain environment based on Chinese Patent Publication No. CN120146357B, fully considers the influence of uncertain factors such as the fluctuation of the processing time of sudden events and the fluctuation of the average driving speed of the inspection vehicle on the planning of the patrol path, as shown in Figure 1 The patrol path optimization method in the present embodiment specifically includes:

[0068] Obtaining basic data of highway patrol in the jurisdiction, real-time traffic operation data and real-time inspection data of highways in the jurisdiction;

[0069] Based on the basic data of highway patrol in the jurisdiction, real-time traffic operation data and real-time inspection data of highways in the jurisdiction, taking the real-time average driving speed of the inspection vehicle and the estimated processing time of the sudden event as uncertain parameters, a patrol scenario set is constructed.

[0070] Taking the minimization of the total cost of the patrol path as the first objective function, the maximization of the total priority value of the to-be-responded events contained in the patrol path as the second objective function, a global dynamic path optimization model is constructed; the minimum total cost of each scenario in the patrol scenario set and the maximum total priority value of the to-be-responded events are obtained through the global dynamic path optimization model. , represents the i-th scenario in the patrol scenario set .

[0071] Taking the minimum total cost expectation value under all scenarios and the maximum total priority expectation value of the to-be-responded events under all scenarios, a patrol path robust optimization model is constructed; based on the and under each scenario, the patrol path robust optimization model is solved, and an optimal robust solution set is output.

[0072] An optimal patrol path is selected from the optimal robust solution set and sent to the inspection vehicle, and the inspection vehicle performs sudden event processing and road section patrol according to the optimal patrol path.

[0073] ​The patrol path optimization method in this embodiment will be described in detail below:

[0074] Preferably, the basic data for highway patrol within the jurisdiction includes: a directed map (RNDG) of the highway network within the jurisdiction, road infrastructure data (BRD), and the prescribed driving speed of the patrol vehicle. Historical patrol time error data (HPTE), all road sections to be patrolled data (ASI), and data on the handling time of each type of historical emergency (HET).

[0075] Specifically, during the entire process of optimizing the patrol route, the basic data of the highway patrol within the jurisdiction only needs to be obtained once. The historical patrol time error data HPTE refers to the deviation between the actual patrol time and the calculation time of the route in the past.

[0076] Preferably, the real-time traffic operation data of the expressway within the jurisdiction includes real-time traffic flow data (RTS).

[0077] Preferably, the real-time inspection data includes: real-time location data (RTPC) of the inspection vehicle, real-time road segment data (RSI) to be inspected, and emergency data (ED). Further, the real-time location data (RTPC) of the inspection vehicle includes: real-time coordinates of the inspection vehicle. The name of the road where the inspection vehicle is located The distance between the inspection vehicle and the highway interchange at the forward end of the road it is located on. (i.e., the distance between the inspection vehicle and the nearest highway interchange in the direction of travel of the road it is on).

[0078] Furthermore, using the real-time average speed of the inspection vehicle and the estimated handling time for emergencies as uncertain parameters, a set of inspection scenarios is constructed. Specifically:

[0079] Select the designated driving speed for the inspection vehicle. and a single emergency Baseline estimated processing time As a baseline scenario This is the number assigned to the emergency.

[0080] Based on real-time traffic flow data (RTS) and historical patrol time error data (HPTE), set A speed less than the prescribed driving speed Fluctuating driving speed The driving speed will be regulated. and A fluctuating driving speed All are used as real-time average driving speeds to construct a speed scenario set. Among them, fluctuating driving speed The set of the fluctuation processing time is represented as , is an adjustment coefficient, is an integer, is an integer greater than or equal to 1;

[0081] According to the emergency event The corresponding historical emergency event type HET and its processing time data HEHT are set fluctuation processing time , fluctuation processing time The fluctuation processing time needs to contain both less than and greater than the reference processing time The reference processing time and fluctuation processing time are all used as the expected processing time to construct the time scenario set ; wherein the set of fluctuation processing time is represented as There are such that and ; , , all represent adjustment coefficients, all are integers, is an integer greater than or equal to 2;

[0082] The speed scenario set and the time scenario set are combined to generate the patrol scenario set , which is further converted into the set , the set and the set The patrol scenarios contained in the set , , represent the patrol scenario composed of the A th real-time average driving speed in the set and the B th expected processing time in the set .

[0083] The person skilled in the art should set the distribution of each fluctuation driving speed according to the real-time traffic flow condition data RTS and the historical patrol time error data HPTE, and set each fluctuation processing time according to the corresponding historical emergency event type HET and its processing time data HEHT of the emergency event; each fluctuation driving speed and each fluctuation processing time Can be set according to experience and actual situation, can be set in advance in the system according to experience The mapping relationship between the real-time traffic flow condition data RTS, the historical patrol time error data HPTE (for example, the real-time traffic flow condition data RTS, the historical patrol time error data HPTE can be respectively matched with a weight, and the fluctuation driving speed The mapping relationship between the sum of the weights of the real-time traffic flow condition data RTS and the historical patrol time error data HPTE), and the fluctuation processing time Corresponding to the same type of historical emergency event processing time data HEHT, in general, the worse the traffic condition reflected by the real-time traffic flow condition data RTS, the greater the historical patrol time error data HPTE, the smaller the lower limit value of the fluctuation driving speed The lower limit value is taken, and the patrol scene set Is constructed in the embodiment to ensure that various scenes that may occur in the patrol process can be covered as much as possible, and the feasibility of the generated patrol path is considered.

[0084] Preferably, since there are various types of emergencies, such as landslides, falling objects, road facility damage, and traffic accidents, the reference expected processing time of each type of emergency is different, so it is necessary to set a time scene set for different types of emergencies. After setting the time scene set of each type of emergency, the corresponding time scene set can be selected for subsequent calculation according to the type of emergency actually occurring in the patrol process. The emergency in the embodiment refers to an event that will affect the normal traffic of the expressway.

[0085] Further, the total cost of the patrol path is minimized As the first objective function, the total priority value of the events to be responded to included in the patrol path is maximized As the second objective function, the global dynamic path optimization model is represented as:

[0086] (1.1),

[0087] (1.2),

[0088] In formulas (1.1) and (1.2): Indicates the total time cost of the patrol vehicle driving and working, Indicates the total penalty cost generated when the patrol vehicle arrives at each event location exceeding the optimal arrival time limit; Indicates the real-time coordinates of the patrol vehicle The starting node All unprocessed emergencies, and a set composed of all to-be-patrolled road sections; S = {e | e is an unprocessed emergency event} S = {e | e is an unprocessed emergency event} S = {e | e is an unprocessed emergency event} S = {e | e is an unprocessed emergency event} S = {e | e is an unprocessed emergency event} S = {e | e is an unprocessed emergency event} S = {e | e is an unprocessed emergency event} S = {e | e is an unprocessed emergency event} S = {e | e is an unprocessed emergency event} S = {e | e is an unprocessed emergency event} S = {e | e is an unprocessed emergency event} S = {e | e is an unprocessed emergency event} S = {e | e is an unprocessed emergency event} S = {e | e is an unprocessed emergency event} S = {e | e is an unprocessed emergency event} S = {e | e is an unprocessed emergency event} S = {e | e is an unprocessed emergency event} S = {e | e is an unprocessed emergency event} S = {e | e is an unprocessed emergency event} S = {e | e is an unprocessed emergency event} S = {e | e is an unprocessed emergency event} S = {e | e is an unprocessed emergency event} S = {e | e is an unprocessed emergency event} S = {e | e is an unprocessed emergency event} S = {e | e is an unprocessed emergency event} S = {e | e is an unprocessed emergency event} S = {e | e is an unprocessed emergency event} S = {e | e is an unprocessed emergency event}

[0089] S = {e | e is an unprocessed emergency event} S = {e | e is an unprocessed emergency event} S = {e | e is an unprocessed emergency event}

[0090] S = {e | e is an unprocessed emergency event} S = {e | e is an unprocessed emergency event} S = {e | e is an unprocessed emergency event}

[0091] S = {e | e is an unprocessed emergency event} S = {e | e is an unprocessed emergency event} S = {e | e is an unprocessed emergency event} S = {e | e is an unprocessed emergency event} S = {e | e is an unprocessed emergency event} S = {e | e is an unprocessed emergency event} S = {e | e is an unprocessed emergency event} S = {e | e is an unprocessed emergency event}

[0092] S = {e | e is an unprocessed emergency event} S = {e | e is an unprocessed emergency event} S = {e | e is an unprocessed emergency event}

[0093] S = {e | e is an unprocessed emergency event} S = {e | e is an unprocessed emergency event} S = {e | e is an unprocessed emergency event}

[0094] In formula (1.4), T is a constant value, representing the expected processing time of the emergency event, T is the time spent on processing the emergency event; The values in different cases are represented as:

[0095] The values in different cases are represented as:

[0096]

[0097] In formula (1.5), T is the time spent on driving the patrol vehicle from the real-time position to the location of the event to be responded, T is the best arrival time limit of the event to be responded; T is the time limit for driving the patrol vehicle from the real-time position to the location of the event to be responded, which is represented as:

[0098]

[0099] In formula (1.6), T is a constant value, representing the best time range from the occurrence of the event to be responded to the arrival of the patrol vehicle for processing, T is the time spent on driving the patrol vehicle in the current time period T when the event to be responded occurs, and T represents the time length between the time when the first emergency event is received by the patrol vehicle and the time when the patrol vehicle stops receiving emergency events.

[0100] Preferably, in the embodiment, the method for distinguishing whether the emergency event belongs to the immediate response event or the event to be responded is as follows: if the emergency event is located on the path between the patrol vehicle and the next unpatrolled path segment in the current patrol path, the emergency event is taken as the immediate response event, and is immediately responded and inserted into the current patrol path; otherwise, the emergency event is taken as the event to be responded and is left for processing.

[0101] Further, the minimum total cost and the maximum total priority value of the events to be responded of the current real-time patrol data in each scenario in the set of patrol scenarios are obtained through the global dynamic path optimization model, specifically:

[0102] The real-time average driving speed and the expected processing time in the i-th scenario in the set of patrol scenarios are extracted, and the and of the current real-time patrol data in the scenario are obtained by using the NSGA-II algorithm (multi-objective genetic algorithm). ​​​​​​​​​​​​​, specifically:

[0103] First, limit the patrol car from its own location, and then according to the priority order of each emergency event to complete the random number of insertion in the emergency event (that is, allow not all emergency events to be inserted, but require the inserted emergency events to meet the priority order), then traverse the to-be-patrolled road segments and each to-be-patrolled road segment is only patrolled once, and finally return the specified node; second, set the population size , the number of chromosome genes , the maximum number of generations , the crossover probability and the mutation probability ; then, initialize the population, perform non-dominated sorting and congestion calculation according to the fitness value of each individual (here, the fitness value is the total cost of the patrol path and the inverse of the total priority value of the to-be-responded events), perform selection, crossover, and mutation operations, and merge the population to perform non-dominated sorting and congestion calculation again to generate a new population; finally, determine whether the termination condition is met, and if so, the iteration ends.

[0104] Specifically, the way in which the NSGA-II algorithm (multi-objective genetic algorithm) solves the global dynamic path optimization model belongs to the common knowledge in the art, and the details not described in detail can be found in Chinese Patent No. CN120146357B and the prior art.

[0105] Further, the patrol path robust optimization model is expressed as:

[0106]

[0107]

[0108] In formulas (1.7) and (1.8), is the set of patrol scenarios, is any scenario in the set of patrol scenarios, is the total number of scenarios in the set of patrol scenarios, is the occurrence probability of scenario , and is the minimum total cost of the feasible solution (i.e., the feasible path scheme) in scenario ; is the maximum total priority value of the to-be-responded events of the feasible solution in scenario , and can be calculated by formulas (1.1) and (1.2), respectively.

[0109] Furthermore, scenario probability constraints and robustness constraints are added to the robust optimization model of the patrol path, specifically:

[0110] The scenario probability constraint is used to limit the sum of the probabilities of all scenarios to 1, and the scenario probability constraint is expressed as:

[0111]

[0112] The robust constraints are used to ensure that the objective function value of the feasible solution in any scenario is close to the optimal objective function value (i.e., ...). and The relative regret value is less than The robust constraints are specifically:

[0113]

[0114]

[0115] In formulas (1.10) and (1.11), The regret factor is... The range of values ​​is ; For the context The minimum total cost; For the context The maximum total priority value of pending events; For a feasible solution (i.e., a feasible path solution) in the scenario The minimum total cost; For a feasible solution in the scenario The total priority value of the highest pending event.

[0116] Furthermore, based on each scenario and Solve the robust optimization model for the patrol path and output the optimal robust solution set, specifically:

[0117] Calculate the minimum total cost under each scenario. set and the total priority value of the maximum pending events set ;in, Let A represent the minimum total cost under the patrol scenario consisting of the A-th real-time average driving speed and the B-th estimated processing time. This represents the total priority value of the maximum pending events in the patrol scenario, which consists of the Ath real-time average driving speed and the Bth estimated processing time.

[0118] The NSGA-II algorithm (multi-objective genetic algorithm) is used to solve the robust optimization model of the inspection path, and the optimal robust solution set is output. And the set of expected total costs corresponding to the optimal robust solution set. and the set of expected total priority values ​​for events to be responded to ;in, Represents the optimal robust solution set The first in A robust solution (i.e., the generated inspection path scheme). express The expected total cost express The expected total priority of events awaiting response;

[0119] Specifically, the NSGA-II algorithm (multi-objective genetic algorithm) solves the robust optimization model for the inspection path as follows:

[0120] A1. Setting Population Size Number of chromosome genes Maximum number of iterations Crossover probability and mutation probability ;

[0121] A2. Generate the initial population ;

[0122] A3. Calculate each individual (i.e., patrol route plan) in the patrol scenario set Total cost and each individual In patrol scenario set Total priority value of pending events ,in, , individual The minimum total cost under the patrol scenario consisting of the A-th real-time average driving speed and the B-th estimated processing time. Represents an individual The maximum total priority value of events awaiting response in the patrol scenario consisting of the Ath real-time average driving speed and the Bth estimated processing time;

[0123] A4. For each individual Determine whether the regret constraint is satisfied: If and Then it is considered that the individual The regret constraint is satisfied at this time. , ,otherwise ;

[0124] in, Represents an individual total cost expectation value of the individual, total priority expectation value of the individual to be responded events in the scenario, probability of the scenario, the first real-time average driving speed and the second predicted processing time, the first real-time average driving speed and the second predicted processing time, the first real-time average driving speed and the second predicted processing time, the first real-time average driving speed and the second predicted processing time, the minimum total cost of the individual in the scenario, the minimum total cost of the individual in the scenario, the maximum total priority value of the individual to be responded events in the scenario;

[0125] A5, the fitness value of the total cost expectation value is , and the fitness value of the total priority expectation value of the to be responded events is ;

[0126] A6, it is judged whether the maximum number of selected generations is reached , if yes, it goes to step A7, if not, the population is sorted and crowdedness calculated according to the fitness value of each individual, the selection, crossover and mutation operations are executed, the population is combined and sorted and crowdedness calculated again to generate a new population , and it re-enters step A3; in the sorting, if or , the solution is dominated solution;

[0127] A7, the current population is taken as the optimal robust solution set, and the optimal robust solution set, the total cost expectation value set corresponding to the optimal robust solution set and the total priority expectation value set of the to be responded events are outputted.

[0128] Further, the optimal patrol path is selected from the optimal robust solution set, which is specifically:

[0129] the total cost expectation value and the total priority expectation value of the to be responded events of each solution in the optimal robust solution set are normalized;

[0130] the total cost expectation value and the total priority expectation value of the to be responded events of each solution after normalization are weighted calculated, since the optimal values of the double objectives are respectively the minimum total cost expectation value and the maximum total priority expectation value of the to be responded events, the smaller the value is, the better the scheme is, and the solution with the smallest value is selected as the optimal patrol path;

[0131] ​​​​

[0132] wherein: is the normalized value of the total cost expectation value corresponding to the solution, is the normalized value of the total priority expectation value of the events to be responded corresponding to the solution, is the weight coefficient of the total cost expectation value, is the weight coefficient of the total priority expectation value of the events to be responded, is the weight coefficient of the total cost expectation value, is the weight coefficient of the total priority expectation value of the events to be responded, The decision maker can value according to his own preference.

[0133] It should be noted that if there is only one solution (i.e., the patrol path scheme) in the optimal robust solution set, the solution is selected as the optimal patrol path.

[0134] The path optimization method of the inspection vehicle in the patrol process in this embodiment is as follows: when there is no sudden event processing demand in the current driving path of the inspection vehicle, real-time sudden events (sudden events can be obtained by multiple sources such as cameras, unmanned aerial vehicle patrol, and alarms) are continuously received, after receiving the first sudden event, other sudden events in a time period T are continuously received, after the time period T ends, all current sudden events and the road segment to be patrolled are combined to perform global path optimization according to the patrol path optimization method of this embodiment, and the generated optimal patrol path is fed back to the remote control end and displayed on the screen of the inspection vehicle; when the inspection vehicle drives according to the optimal patrol path and processes all the sudden events, the inspection vehicle starts to receive sudden events again, and the cycle continues until the patrol task of all road segments is completed. For matters not fully described in this embodiment, please refer to Chinese Patent No. CN120146357B and the prior art.

[0135] The patrol path optimization method of this embodiment uses robust optimization theory, considers that the actual average driving speed of the staff during the patrol process will be lower than the specified vehicle speed due to random events along the way, and that the actual processing time of a single sudden event has uncertainty, takes the real-time average driving speed of the inspection vehicle and the estimated processing time of a single sudden event as uncertain parameters, and clearly distinguishes the fluctuation scenarios of the real-time average driving speed of the inspection vehicle and the estimated processing time of a single sudden event to ensure that various scenarios that may occur during the patrol process are covered as much as possible. At the same time, on the basis of the global dynamic path optimization model, a patrol path robust optimization model is further constructed to minimize the total cost expectation value under all scenarios and maximize the total priority expectation value of the events to be responded under all scenarios, and the optimal patrol path under the current real-time inspection data (i.e., the current road segment to be patrolled and the sudden events) is solved through the patrol path robust optimization model, thereby improving the feasibility of the patrol path and taking into account the requirements of sudden event processing and completing all road segment inspections.

[0136] Embodiment 2:

[0137] This embodiment is based on part of the highways around Changsha City, and the highway patrol path planning is carried out by using the highway patrol path optimization method in example 1 under uncertain environment, as follows:

[0138] Figure 2 For the directed graph RNDG of the selected highway network, one highway interchange is taken as one node, and the nodes are numbered as 1-8, a total of 8 nodes and 20 edges are included, and the weight on each edge represents the length (km) of the corresponding highway section, wherein node 4 is taken as the starting node.

[0139] As shown in Figure 3 , o in the figure represents the current position of the patrol vehicle, R1, R2, R3, and R4 are to-be-patrolled road sections, and the current remaining patrol path of the patrol vehicle is o→R2→5→R1→6→8→R4→7→R3→8→6→4. At this time, the patrol vehicle receives 3 emergency event processing demands, and the information and position distribution of each emergency event are shown in Table 1 and Figure 4 .

[0140] Table 1 Emergency event information table

[0141]

[0142] Further, in this embodiment, the specified driving speed of the patrol vehicle is taken as 80 km / h, the single-emergency-event reference expected processing time is taken as shown in Table 1; 2 fluctuation driving speeds less than the specified driving speed are set , to form a driving speed scenario set ; 1 fluctuation processing time less than and 1 fluctuation processing time greater than the single-emergency-event reference expected processing time are set , to form an expected processing time scenario set ; the speed scenario set and the time scenario set are combined to generate a patrol scenario set containing 9 patrol scenarios , which is further converted into set , set , and . The patrol scenarios contained in set , set , and are unchanged, and each patrol scenario is shown in Table 2: Table 2 Patrol scenario set

[0143]

[0144]

[0145] ​​The real-time average driving speed and the expected processing time in each patrol scene are extracted, and the NSGA-II algorithm is used to solve the minimum total cost in each patrol scene through a global dynamic path optimization model and the maximum total priority value of events to be responded As shown in Table 3:

[0146] Table 3: Patrol path under each patrol scene and

[0147]

[0148] Further, the regret coefficient = 0.5 is set in the embodiment, based on the and The NSGA-II algorithm is used to solve the robust optimization model to obtain a unique final path scheme o→5→ →7→ →8→ →7→5→ →6→ →5→3→ The total cost of the path is 3.13h, and the total priority value of the events to be responded is 1.0, and the path is shown in Figure 5 .

[0149] Figure 6 The patrol path generated by the Chinese patent with publication number CN120146357B under the same condition is shown in the figure, and the patrol path is o→5→ →7→A2→5→7→ →8→ →7→5→ →6→ →5→3→ Compared with the Chinese patent with publication number CN120146357B, the patrol path scheme of the embodiment adopts a more conservative way to only process the emergency event A1, which can take into account the emergency event processing and complete the scheduled road section patrol task.

[0150] Embodiment 3:

[0151] The embodiment provides a highway patrol path optimization system in an uncertain environment, comprising:

[0152] A data acquisition and storage module is configured to acquire and store basic data of highway patrol in a jurisdiction, real-time traffic operation data of highways in the jurisdiction, real-time inspection data, and a regret coefficient.

[0153] The patrol scene generation module constructs a patrol scene set taking the real-time average driving speed of the patrol vehicle and the emergency event estimated processing time as uncertain parameters ;

[0154] The global dynamic path optimization model solution module is configured to obtain the minimum total cost and the maximum total priority value of events to be responded to in each scenario and the maximum total priority value of events to be responded to in each scenario ;

[0155] The patrol path robust optimization model solution module is configured to call the global dynamic path optimization model solution module to obtain the minimum total cost and the maximum total priority value of events to be responded to in each scenario corresponding to each feasible solution, to output an optimal robust solution set with the minimum total cost in all scenarios and the maximum total priority value of events to be responded to in all scenarios as a target, and to filter out an optimal patrol path.

[0156] The information feedback module is configured to feed back the optimal patrol path to the remote control end and display the optimal patrol path on a screen of the patrol vehicle.

[0157] Embodiment 4

[0158] The embodiment provides a highway patrol path optimization system in an uncertain environment, including a memory and a processor, the memory stores a computer program, and the processor executes the computer program to perform the highway patrol path optimization method in the uncertain environment in the embodiment 1.

[0159] Embodiment 5

[0160] The embodiment provides a storage medium, the storage medium stores a computer program, and the computer program is executed to perform the highway patrol path optimization method in the uncertain environment in the embodiment 1.

[0161] The above merely describes the preferred embodiments of the present application and is not used to limit the present application. For those skilled in the art, the present application can have various modifications and changes. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A method for optimizing a highway patrol path in an uncertain environment, characterized in that, include: Obtain basic data on highway patrols within the jurisdiction, real-time traffic operation data and real-time inspection data for highways within the jurisdiction; With the real-time average driving speed of the inspection vehicle and the estimated processing time of the emergency as uncertain parameters, a patrol scene set is constructed ; To minimize the total cost of the inspection path The first objective function is to maximize the total priority value of the events to be responded to contained in the inspection path. For the second objective function, a global dynamic path optimization model is constructed; the patrol scenario set is obtained through the global dynamic path optimization model. Minimum total cost under various scenarios and the total priority value of the maximum pending events , Represents a set of patrol scenarios The first in A scenario; A robust optimization model of the patrol path is constructed with the minimum total cost expectation value in all scenarios and the maximum total priority expectation value of the events to be responded in all scenarios. and The robust optimization model of the patrol path is solved, and an optimal robust solution set is output. Select the optimal inspection path from the set of optimal robust solutions and send it to the inspection vehicle; The real-time average driving speed of the inspection vehicle and the estimated processing time of the emergency event are taken as uncertain parameters to construct a patrol scene set , and specifically: Selecting a patrol vehicle prescribed travel speed and single incident baseline expected processing time as a baseline scenario; Setting a fluctuation driving speed less than a prescribed driving speed , the prescribed driving speed and the fluctuation driving speed are all taken as real-time average driving speeds to construct a speed scenario set ; wherein the set of fluctuation driving speeds is expressed as , is an adjustment coefficient, is an integer, is an integer greater than or equal to 1; Setting a fluctuation processing time , a reference processing time and a fluctuation processing time are all taken as a set of time scenarios of a processing time ; wherein the set of fluctuation processing times is expressed as , there are such that and ; , , all represent adjustment coefficients, all are integers, is an integer greater than or equal to 2; combining a set of speed scenarios and a set of time scenarios to generate a set of patrol scenarios further transforming into a set a set and a set the patrol scenarios contained in the set , , denotes a patrol scenario composed of the A-th real-time average travel speed in the set and the B-th estimated processing time in the set .

2. The method of claim 1, wherein, The robust optimization model for the patrol path is expressed as follows: In formulas (1.7) and (1.8), for a set of patrol scenarios, for any scenario in the set of patrol scenarios, for a scenario the probability of occurrence of, the minimum total cost of a feasible solution under scenario ; the maximum total priority value of events to be responded to under scenario .

3. The method of claim 2, wherein, The scenario probability constraints of the robust optimization model for the patrol path are expressed as follows: The robust constraints of the robust optimization model for the patrol path are expressed as follows: In formulas (1.10) and (1.11), is a regret coefficient, is in the range .

4. The method of claim 1, wherein, Based on the scenarios and Solving the robust optimization model of the patrol path, outputting the optimal robust solution set, specifically: Calculate the minimum total cost under each scenario. set and the total priority value of the maximum pending events set ;in, Let A represent the minimum total cost under the patrol scenario consisting of the A-th real-time average driving speed and the B-th estimated processing time. This represents the total priority value of the maximum pending events in the patrol scenario, which consists of the Ath real-time average driving speed and the Bth estimated processing time. Solve the robust optimization model of the patrol path using the NSGA-II algorithm, output the optimal robust solution set , and the total cost expectation value set corresponding to the optimal robust solution set , and the total priority expectation value set of the events to be responded ; wherein, represents the th robust solution in the optimal robust solution set , represents the total cost expectation value of , represents the total priority expectation value of the events to be responded.

5. The method of claim 4, wherein, The NSGA-II algorithm solves the robust optimization model for inspection paths as follows: A1, set population size , chromosome gene number , maximum generation number , crossover probability and mutation probability ; A2. Generating initial population ; A3, the total cost of each individual under the set of patrol scenarios and each individual under the set of patrol scenarios the total priority value of pending events of each individual under the set of patrol scenarios ; wherein, , the minimum total cost of each individual under the patrol scenario consisting of the A-th real-time average travel speed and the B-th estimated processing time, the maximum total priority value of pending events of each individual under the patrol scenario consisting of the A-th real-time average travel speed and the B-th estimated processing time; A4, for each individual whether the regret constraint is satisfied is determined: if and then the individual is considered to satisfy the regret constraint, in which case , otherwise ; wherein, represents the total cost expectation of the individual , represents the total priority expectation of the individual to respond to the event, is the probability of the scenario occurring, refers to a patrol scenario composed of the first real-time average travel speed and the second estimated processing time, represents the minimum total cost of the individual under the scenario , represents the maximum total priority value of the individual to respond to the event under the scenario ; A5, the fitness value of the total cost expectation value is , the fitness value of the total priority expectation value of the events to be responded to is ; A6. Determine if the maximum number of generations has been reached. If the fitness value is reached, proceed to step A7; otherwise, adjust the population based on the fitness value of each individual. Perform non-dominated sorting and crowding calculation, execute selection, crossover, and mutation operations, merge the populations, and then perform non-dominated sorting and crowding calculation again to generate a new population. And re-enter step A3; in non-dominated sorting, if there exists or Then let the solution be found. For the subordinating solution; A7. Using the current population as the optimal robust solution set, output the optimal robust solution set, the expected total cost set corresponding to the optimal robust solution set, and the expected total priority set of the events to be responded to.

6. The method of claim 1, wherein, When there are two or more solutions in the optimal robust solution set, the specific steps for selecting the optimal inspection path are as follows: The expected total cost of each solution and the expected total priority of the events to be responded to in the optimal robust solution set are normalized. For each solution, the normalized expected total cost and the expected total priority of the events to be responded to are weighted and calculated, and then selected. The solution with the smallest value is taken as the optimal inspection path; wherein: is a normalized value of the total cost expectation value corresponding to the solution, is a normalized value of the total priority expectation value of the events to be responded corresponding to the solution, is a normalized value of the total cost expectation value corresponding to the solution, is a normalized value of the total priority expectation value of the events to be responded corresponding to the solution, is a weight coefficient of the total cost expectation value, is a weight coefficient of the total priority expectation value of the events to be responded, ; If there is only one solution in the optimal robust solution set, then that solution is selected as the optimal inspection path.

7. A highway patrol path optimization system in an uncertain environment, characterized by, The system employs the patrol path optimization method as described in any one of claims 1-6, and the system includes: The data acquisition and storage module is used to acquire and store basic data of highway patrol within the jurisdiction, real-time traffic operation data of highways within the jurisdiction, real-time inspection data, and regret coefficient. The patrol scene generation module takes the real-time average driving speed of the patrol vehicle and the emergency event estimated processing time as uncertain parameters to construct a patrol scene set ; The global dynamic path optimization model solving module is configured to obtain the minimum total cost under each scenario and the maximum total priority value of events to be responded ; The patrol path robust optimization model solution module calls the global dynamic path optimization model solution module to obtain the minimum total cost and the maximum total priority value of the pending events corresponding to each feasible solution in each scenario. With the goal of minimizing the expected total cost value in all scenarios and maximizing the expected total priority value of the pending events in all scenarios, it outputs the optimal robust solution set and selects the optimal patrol path. The information feedback module is used to feed back the optimal patrol route to the remote control terminal and display it on the screen of the patrol vehicle.

8. A highway patrol path optimization system in an uncertain environment, characterized in that, It includes a memory and a processor, wherein the memory stores a computer program, and the processor executes the patrol path optimization method as described in any one of claims 1-6 when running the computer program.

9. A storage medium, characterized by The storage medium stores a computer program, which, when run, executes the patrol path optimization method as described in any one of claims 1-6.

Images