Path determination method and device, storage medium, electronic equipment and program product

Abstract

The invention discloses a path determination method and device, a storage medium, electronic equipment and a program product, and relates to the field of financial science and technology. The method comprises the steps that a route map is determined according to the position of a cash truck and the positions of N network points, and the route map at least comprises path information and traffic information from the cash truck to each network point and path information and traffic information between any two network points in the N network points; acquiring a set priority of each branch for the cash transport task; and determining a target driving path of the cash truck according to the route map, the set priority of each branch for the cash transport task and the vehicle state data of the cash truck. The technical problem that the driving path planning accuracy of the cash truck is low due to the fact that traditional cash truck path planning depends on a fixed route or simple artificial experience and is difficult to cope with a complex and changeable traffic environment is solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of financial technology, in particular to a path determination method and device, a storage medium, an electronic device and a program product. BACKGROUND

[0002] As a key link in financial logistics, the cash van undertakes the transportation task of high-value cash, valuable goods and other sensitive materials. Its transportation process not only involves the safe transfer of huge funds, but also directly relates to the normal operation of financial institutions and the stable operation of the financial system. Therefore, the safety and transportation efficiency of the cash van have always been an important concern of financial logistics.

[0003] In traditional cash van path planning, fixed routes or simple manual experience are usually relied on. Although these methods can meet the basic transportation needs to some extent, their limitations gradually become apparent when faced with complex and variable traffic environments and potential security threats. For example, traditional path planning often uses fixed preset routes, which are rigid and difficult to adapt to changes in traffic conditions in real time. Fixed routes can cause the cash van to be trapped in congestion or delayed, and even increase security risks. When relying on manual experience for path planning, the judgment of the planning personnel is often influenced by subjective factors, and the accuracy and reliability of path planning are low. SUMMARY

[0004] The embodiments of the present application provide a path determination method, device, storage medium, electronic device and program product to at least solve the technical problem that traditional cash van path planning relies on fixed routes or simple manual experience, is difficult to cope with complex and variable traffic environments, and results in low accuracy of cash van driving path planning.

[0005] According to an aspect of an embodiment of the present application, a path determination method is provided, comprising: determining a route graph according to the location of a cash van and the location of N network points, wherein N is an integer greater than 1, and the route graph at least includes: path information and traffic information from the cash van to each network point, and path information and traffic information between any two network points in the N network points; obtaining a set priority of each network point for a cash transportation task; determining a target driving path of the cash van according to the route graph, the set priority of each network point for the cash transportation task and vehicle state data of the cash van, wherein the target driving path includes the N network points.

[0006] Optionally, the route map is determined according to the location of the cash transport vehicle and the locations of the N network points, including: obtaining geographical position information of target locations, wherein the target locations are locations with a crowd density greater than a preset threshold and / or locations with a frequency of public security events greater than a preset frequency; marking the geographical position information of the target locations in the route map, and establishing a path connection relationship between the target locations and the location of the cash transport vehicle and the locations of each network point.

[0007] Optionally, the target driving path of the cash transport vehicle is determined according to the route map, the set priority of each network point for the cash transport task, and the vehicle state data of the cash transport vehicle, including: collecting route planning reference information, wherein the route planning reference information at least includes: meteorological data, historical driving routes of the cash transport vehicle when performing historical cash transport tasks, and time consumption information based on the historical driving routes to reach each network point; determining the target driving path of the cash transport vehicle according to the route planning reference information, the route map, the set priority of each network point for the cash transport task, and the vehicle state data of the cash transport vehicle.

[0008] Optionally, the target driving path of the cash transport vehicle is determined according to the route planning reference information, the route map, the set priority of each network point for the cash transport task, and the vehicle state data of the cash transport vehicle, including: constructing a target feature matrix according to the route planning reference information, the route map, the set priority of each network point for the cash transport task, and the vehicle state data of the cash transport vehicle; inputting the target feature matrix into a target model, and determining the target driving path of the cash transport vehicle through prior knowledge learned by the target model in a training process.

[0009] Optionally, the target driving path of the cash transport vehicle is determined by inputting the target feature matrix into a target model and through prior knowledge learned by the target model in a training process, including: determining a driving environment of the cash transport vehicle by the target model according to the target feature matrix; taking a driving action of the cash transport vehicle as an operation behavior in the driving environment; determining a target driving path of the cash transport vehicle according to the driving environment and the operation behavior through prior knowledge of the target model, wherein the target model learns target operations of the cash transport vehicle in different driving environment states in a training process, and the target operations ensure that the cash transport vehicle travels through the N network points in the shortest time.

[0010] Optionally, the training process of the target model includes the following steps:

[0011] obtaining training samples, wherein the training samples include M reference routes formulated based on the route map, and M is an integer greater than 1;

[0012] According to the training sample, the first constraint condition and the second constraint condition, the neural network is iteratively trained until the neural network enters a convergent state, and the target model is trained, wherein the first constraint condition is used to constrain the neural network to design a new route according to the M reference routes, and the design target is to design a route with the shortest time consumption under the condition that the set priority of each network point for the cash transport task is met; the second constraint condition is used to set a positive reward for reaching a network point in advance and a negative reward for reaching a network point overtime in the process of designing a route.

[0013] Optionally, according to the route planning reference information, the route graph, the set priority of each network point for the cash transport task and the vehicle state data of the cash transport vehicle, a target feature matrix is constructed, including: creating an initial feature matrix, wherein the rows of the initial feature matrix represent each road location in the route graph and the N network points, and the columns represent the feature information to be filled; filling corresponding data for each cell of the initial feature matrix to obtain the target feature matrix, wherein for each road node in the route graph, the average speed, congestion index and weather influence coefficient of the cash transport vehicle when driving to the road node are filled; for each network point, the set priority, predicted cash transport vehicle arrival time and probability of occurrence of a public security event of the network point are filled.

[0014] According to another aspect of the present application, a path determination apparatus is also provided, wherein the apparatus comprises: a first determination unit configured to determine a route graph according to a location of a cash transport vehicle and network point locations of N network points, wherein N is an integer greater than 1, and the route graph at least comprises path information and traffic information from the cash transport vehicle to each network point and path information and traffic information between any two network points in the N network points; a first acquisition unit configured to acquire a set priority of each network point for a cash transport task; and a second determination unit configured to determine a target driving path of the cash transport vehicle according to the route graph, the set priority of each network point for the cash transport task and vehicle state data of the cash transport vehicle, wherein the target driving path comprises the N network points.

[0015] According to another aspect of the present application, a computer readable storage medium is also provided, wherein the computer readable storage medium stores a computer program, and when the computer program runs, the computer readable storage medium causes a device where the computer readable storage medium is located to perform the path determination method described above.

[0016] According to another aspect of the present application, an electronic device is also provided, wherein the electronic device comprises one or more processors and a memory for storing one or more programs, wherein the one or more programs, when executed by the one or more processors, cause the one or more processors to perform the path determination method described above.

[0017] According to another aspect of the present application, a computer program product is also provided, wherein the computer program product comprises computer programs or instructions which, when executed by a processor, implement the path determination method described above.

[0018] In the present application, first, a route map is determined according to the location of the cash transport vehicle and the locations of N network points, wherein N is an integer greater than 1, and the route map at least comprises path information and traffic information from the cash transport vehicle to each network point, and path information and traffic information between any two network points of the N network points. Then, the set priority of each network point for the cash transport task is obtained, and a target driving path of the cash transport vehicle is determined according to the route map, the set priority of each network point for the cash transport task, and vehicle state data of the cash transport vehicle, wherein the target driving path includes the N network points.

[0019] From the above, it can be seen that, unlike traditional extensive path selection, the present application realizes the refinement and personalization of path planning by collecting and analyzing detailed path information and traffic information from the cash transport vehicle to each network point, and the connection between any two network points, ensuring that each path is a decision based on the currently optimized data. By introducing the concept of network point priority, path planning not only considers distance and time, but also considers the urgency and importance of the task. This priority-based path optimization mechanism greatly improves the timeliness and efficiency of the cash transport task, especially in the face of emergencies, it can prioritize the completion of critical tasks.

[0020] In addition, in the present application, the vehicle state data of the cash transport vehicle is included in the consideration range of path planning, which means that the system can develop more reasonable driving strategies according to the actual conditions of the vehicle (such as fuel quantity, maintenance needs, etc.), avoiding task failure or safety problems caused by poor vehicle state.

[0021] The comprehensive application of the above-mentioned technologies not only improves the efficiency and accuracy of a single cash transport task, but also promotes the optimized scheduling of the cash transport vehicle in the entire service network, reduces invalid driving and waiting time, and thus improves the operation efficiency of the entire cash transport vehicle service system.

[0022] In summary, the technical scheme of the present application overcomes the limitations of traditional path planning by combining dynamic route map with priority, and effectively overcomes the limitations of traditional path planning by combining real-time state monitoring of the cash transport vehicle, thereby achieving efficient and safe execution of the cash transport task. BRIEF DESCRIPTION OF DRAWINGS

[0023] The accompanying drawings, which are included to provide a further understanding of the present application and are incorporated in and constitute a part of this application, illustrate embodiments of the present application and together with the description serve to explain the present application. In the drawings:

[0024] Figure 1 is a flow chart of an optional path determination method according to an embodiment of the present application;

[0025] Figure 2 is a training flow chart of an optional target model according to an embodiment of the present application;

[0026] Figure 3 is a schematic diagram of an optional path determination device according to an embodiment of the present application. DETAILED DESCRIPTION

[0027] In order to make the personnel in the technical field better understand the present application, the technical scheme in the embodiments of the present application will be described clearly and completely below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, but not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor should belong to the scope of protection of the present application.

[0028] It should be noted that the terms "first", "second", and the like in the specification and claims of the present application and the above-described drawings are used to distinguish similar objects, and do not necessarily indicate a specific order or a chronological sequence. It should be understood that the data thus used can be interchanged under appropriate circumstances, so that the embodiments of the present application described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion, for example, a process, method, system, product or device that includes a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but can include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0029] It should be noted that the information (including but not limited to user equipment information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) collected by the present application are information and data authorized by the user or authorized by all parties, and the collection, storage, use, processing, transmission, provision, disclosure and application of related data comply with relevant laws, regulations and standards in the relevant region, necessary security measures are taken, and the public order is not violated, and appropriate operation portal is provided for user to choose authorization or refusal. For example, an interface is provided between the system and the related user or institution, and before obtaining the relevant information, the interface needs to send a request to the aforementioned user or institution, and after receiving the consent information feedback from the aforementioned user or institution, the relevant information is obtained.

[0030] According to the embodiment of the present application, a cash transport vehicle navigation system can be used as the execution subject of the path determination method of the embodiment of the present application, wherein the system can be a software system or an embedded system combining software and hardware. Of course, the method execution subject in the embodiment of the present application can also be other forms of execution subject, such as device, equipment, etc. Those skilled in the art should know that the present application does not particularly limit the specific form of the method execution subject.

[0031] According to the embodiment of the present application, a path determination method embodiment is provided. It should be noted that the steps shown in the flowchart of the drawing can be executed in a computer system such as a group of computer executable instructions, and although the logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that here.

[0032] Figure 1 According to the embodiment of the present application, an optional path determination method is shown in the flowchart of the present application, as shown in Figure 1 The method comprises the following steps:

[0033] Step S101, determining a route map according to the location of the cash transport vehicle and the location of N network points, wherein N is an integer greater than 1, and the route map at least includes: path information and traffic information from the cash transport vehicle to each network point, and path information and traffic information between any two network points in the N network points.

[0034] In step S101, the location of the cash transport vehicle refers to the real-time geographic position of the cash transport vehicle when performing the task, which can be obtained by satellite navigation system or other positioning technology, and is used as the starting point of dynamic path planning. The location of the N network points refers to the set of financial nodes (such as bank network points, ATM machines, vaults, etc.) that need to be served in the cash transport task, N is the number of network points, and at least contains two network points to ensure the necessity of path planning.

[0035] Optionally, the route graph can be a structured data model for describing the path relationship between the cash transport vehicle and each site. The path information can refer to static road network data (such as distance, road level, speed limit, etc.), and the traffic information can refer to dynamic traffic data (such as real-time congestion index, traffic accident, construction road closure, etc.).

[0036] It should be noted that the route graph at least includes: path information and traffic information from the cash transport vehicle to each site, and path information and traffic information between any two sites of the N sites. By integrating real-time traffic information, the route graph can be dynamically updated to avoid congested or dangerous sections, significantly improving the accuracy of path planning. In addition, the route graph not only contains the path from the cash transport vehicle to a single site, but also covers the path relationship between the N sites, supporting the optimal ordering of multi-site tasks and reducing the total travel time.

[0037] In step S102, the set priority of each site for the cash transport task is obtained.

[0038] In step S102, the set priority of each site for the cash transport task will be different. For example, some sites can be high-value sites, such as large daily cash volume, which need to be serviced first to avoid bank liquidity risk due to delay. Or some sites are time-sensitive nodes, such as ATM pre-operating cash replenishment, which have strict time window constraints, and overtime will trigger regulatory penalties. By quantifying the priority, the resource allocation to key nodes is ensured.

[0039] In step S103, a target travel path of the cash transport vehicle is determined according to the route graph, the set priority of each site for the cash transport task, and the vehicle state data of the cash transport vehicle, wherein the target travel path includes N sites.

[0040] In step S103, the route graph is the basis of intelligent navigation, providing road network topology, which can be updated in real time. The set priority of each site for the cash transport task is a quantitative expression of the site business demand, which allocates decision weights. The vehicle state data of the cash transport vehicle is a hard boundary of physical constraints, which can include oil or electricity endurance, vehicle position, speed, fuel consumption, or bulletproof cabin door lock state, so that the path planning can be optimized according to the actual status of the vehicle, for example, to avoid selecting a long route when the vehicle has low oil level.

[0041] From the above, by introducing the route map, the priority of each node for the cash transport task, and the dynamic collection of multi-source data such as the vehicle state data of the cash transport vehicle, the target driving path of the cash transport vehicle is determined, which covers all aspects of the cash transport task, can more comprehensively reflect the current transportation environment, and dynamically adjusts according to the current actual situation, reduces the dependence on fixed routes or simple manual experience, can better cope with complex and variable traffic environment, and improves the accuracy of the driving path planning of the cash transport vehicle.

[0042] In an optional embodiment, the route map is determined according to the location of the cash transport vehicle and the location of the N nodes, comprising: obtaining the geographic location information of the target location through the cash transport vehicle navigation system, wherein the target location is a location with a crowd density greater than a preset threshold and / or a location with a frequency of public security events greater than a preset frequency. Then the cash transport vehicle navigation system labels the geographic location information of the target location in the route map, and establishes a path connection relationship between the target location and the location of the cash transport vehicle and the location of each node.

[0043] Optionally, the target location is a specific location that needs special attention in the cash transport path planning. The target location can be a location with a crowd density greater than a preset threshold, which may have a high security risk due to the dense crowd, such as shopping malls, transportation hubs, etc. The cash transport vehicle needs to pay special attention to these areas when planning the path, and may need to avoid or take additional safety measures.

[0044] The target location can also be a location with a frequency of public security events greater than a preset frequency, which is considered a high-risk area due to the high frequency of public security events. The cash transport vehicle also needs to avoid these locations as much as possible when planning the path to reduce the security risk.

[0045] The geographic location information of the above-mentioned target location refers to the specific location data of the target location, including latitude and longitude coordinates, address, etc. The path connection relationship refers to the connection path between the current location of the cash transport vehicle, the target location and each node labeled in the route map. These paths not only include physical road connections, but also may include traffic information to help the system calculate the optimal path.

[0046] It should be noted that the cash transport vehicle navigation system can obtain the geographic location information of the target location through geographic information system data, public security event database or other related data sources. The geographic location information of the target location will be used for subsequent path planning and risk assessment, and the obtained geographic location information of the target location will be labeled in the route map constructed before, and the purpose of labeling is to enable the system to intuitively see the location of these high-risk areas, and to consider these factors when planning the path.

[0047] Optionally, the target driving path of the cash transport vehicle is determined according to the route map, the set priority of each node for the cash transport task, and the vehicle state data of the cash transport vehicle, including: collecting route planning reference information through the cash transport vehicle navigation system, wherein the route planning reference information at least includes: meteorological data, historical driving routes of the cash transport vehicle when performing historical cash transport tasks, and time consumption information based on the historical driving routes to reach each node. The cash transport vehicle navigation system can also determine the target driving path of the cash transport vehicle according to the route planning reference information, the route map, the set priority of each node for the cash transport task, and the vehicle state data of the cash transport vehicle.

[0048] Optionally, the route planning reference information described above is used as additional data for auxiliary path planning. For example, meteorological data includes weather conditions and their impact on road traffic. Adverse weather conditions can increase travel time or risk, so they need to be considered when planning the path. Historical driving routes refer to the actual driving routes of the cash transport vehicle when performing historical cash transport tasks. These data can provide experience reference to help the cash transport vehicle navigation system understand which routes perform better in actual operation. Time consumption information refers to the actual time consumption based on the historical driving routes to reach each node. These data reflect the time efficiency of different routes in actual operation, which helps to optimize path planning.

[0049] It should be noted that by collecting route planning reference information, the cash transport vehicle navigation system can integrate route planning reference information, route map, set priority of each node for the cash transport task, and vehicle state data of the cash transport vehicle to form a complete data set. The integrated data can extract features related to path planning, so that not only traffic and meteorological data are considered, but also historical driving routes and time consumption information are combined, and then more reasonable decisions can be made when planning the path.

[0050] Optionally, the target driving path of the cash transport vehicle is determined according to the route planning reference information, the route map, the set priority of each node for the cash transport task, and the vehicle state data of the cash transport vehicle, including: the cash transport vehicle navigation system can construct a target feature matrix according to the route planning reference information, the route map, the set priority of each node for the cash transport task, and the vehicle state data of the cash transport vehicle. Then, the cash transport vehicle navigation system can input the target feature matrix into the target model to determine the target driving path of the cash transport vehicle through the prior knowledge learned by the target model in the training process.

[0051] Optionally, the target feature matrix is a data structure that integrates and organizes various feature data related to the armored car route planning. The target feature matrix can be a table-form data, where each row represents a data point, such as a road node or a grid point, and each column represents a feature, such as average speed, congestion index, priority, etc. This matrix can be used as input for the route planning algorithm, for model training and prediction.

[0052] Optionally, the target model can be a machine learning or deep learning-based algorithm model that predicts the optimal driving path of the armored car based on the input feature matrix. The target model learns the patterns and rules of the armored car driving environment during the training process and can generate the optimal path based on new input data. The target model can be a deep learning network, random forest, support vector machine, etc.

[0053] Optionally, the prior knowledge of the target model refers to the experience and rules learned by the target model from historical data during the training process. These knowledge includes the optimal driving strategy of the armored car in different environments, the safety evaluation of different paths, and how to adjust the path according to the grid priority and vehicle state, etc. Prior knowledge enables the model to make accurate predictions on new input data.

[0054] In the embodiments of the present application, the armored car navigation system can integrate the route planning reference information, the route map, the set priority of each grid point for the armored car task, and the vehicle state data of the armored car, construct a target feature matrix, and input the constructed target feature matrix into the target model. The target model is a trained machine learning or deep learning model that can process the input feature data. The target model uses its prior knowledge learned during the training process to analyze and process the input feature matrix and generate the optimal driving path. The path output by the target model is the optimal path after considering all factors, i.e., the target driving path. The armored car navigation system can automatically learn and adapt to complex traffic environments and business demands through the machine learning model to generate intelligent path planning results. The machine learning model learns a large amount of historical data and experience rules during the training process, which can more accurately predict the optimal path, and has higher accuracy compared to traditional manual experience or fixed route methods. With continuous input of new data and continuous training of the model, the target model can continuously optimize its prior knowledge and further improve the performance of path planning.

[0055] Optionally, the target feature matrix is ​​input into the target model, and the target driving path of the armored car is determined through the prior knowledge learned by the target model during the training process, including: determining the driving environment of the armored car based on the target feature matrix through the target model; using the driving action of the armored car as the operating behavior in the driving environment; determining the target driving path of the armored car based on the driving environment and the operating behavior through the prior knowledge of the target model, wherein the target model learns the target operation of the armored car under different driving environment conditions during the training process, wherein the target operation ensures that the armored car passes through N outlets in the shortest time.

[0056] Alternatively, the driving environment can be the external conditions faced by the armored car when performing its mission, including traffic conditions, weather conditions, public security conditions, road network structure, etc. The driving environment is described by the feature data in the target feature matrix. Driving actions refer to the operational behaviors that the armored car can take during driving, such as acceleration, deceleration, turning, and stopping. In path planning, driving actions can be abstracted as the behavior of selecting a specific path or node. The target operation refers to the optimal driving strategy recommended by the target model based on the driving environment and driving actions. The goal of the target operation is to safely pass through all network points in the shortest time while meeting time constraints and safety requirements.

[0057] In an embodiment of the present application, the armored car navigation system can input a target feature matrix into a target model. The target feature matrix contains various features of the armored car's driving environment, such as road congestion, weather conditions, network priority, etc. The target model identifies the current driving environment based on the input feature matrix. For example, the model can determine whether the current traffic conditions are congested, whether the weather is bad, whether it is in a high-risk area, etc. Driving actions are defined as operational behaviors that the armored car can take in the driving environment. For example, choosing a certain path, turning at a certain node, accelerating or decelerating, etc. The target model maps the driving actions to the current driving environment and analyzes the impact of these operational behaviors on the driving path. For example, choosing a path with less congestion can reduce driving time.

[0058] The target model leverages the prior knowledge learned during training to predict the optimal driving path based on the current driving environment and maneuvers. By analyzing the driving environment and operational behavior, the target model generates a target driving path that satisfies the requirements, allowing the armored vehicle to safely pass through all locations in the shortest possible time. By learning from prior knowledge, the target model intelligently generates the optimal path based on the current driving environment and maneuvers. This is more accurate and efficient than traditional manual experience, and the target model can analyze changes in the driving environment in real time and dynamically adjust the route plan.

[0059] Optionally, Figure 2is a flowchart of an optional training process of a target model according to an embodiment of the present application, wherein, as shown in Figure 2 the training process of the target model includes the following steps:

[0060] In step S201, a training sample is obtained, wherein the training sample includes M reference routes formulated based on a route atlas, and M is an integer greater than 1.

[0061] In step S202, the neural network is iteratively trained according to the training sample, a first constraint condition and a second constraint condition until the neural network enters a convergent state, and a target model is trained.

[0062] The first constraint condition is used to constrain the neural network to design a new route based on the M reference routes, and the design goal is to design a route with the shortest time consumption under the condition that the set priority of each network point for the cash transport task is met; the second constraint condition is used to set a positive reward for reaching a network point in advance and a negative reward for reaching a network point overtime during the design of the route.

[0063] Optionally, the training sample is a data set used to train the target model, which contains multiple reference routes formulated based on a route atlas. These reference routes are designed according to historical data or experience, and are used to guide the model to learn how to generate the optimal path. The reference route refers to the example path provided to the model during the training process. These paths are the actual driving routes from the starting position of the cash transport vehicle to each network point, which are usually extracted from historical cash transport tasks or designed by experts according to experience.

[0064] Optionally, the neural network is a machine learning-based algorithm model that simulates the connection mode of human brain neurons to process and analyze data. Here, the neural network is used to learn the rules and patterns of cash transport vehicle path planning. The convergent state refers to the state in which the output of the neural network gradually stabilizes during the training process, and the error no longer significantly decreases. When the neural network enters the convergent state, it means that the model has learned enough rules and can be used for actual path planning tasks.

[0065] It should be noted that the first constraint condition described above is one of the constraints on the neural network during the training process, which requires the model to design a new route based on the reference route. At the same time, the goal is to generate a route with the shortest time consumption under the premise of meeting the priority of each network point. This ensures that the path generated by the model not only meets the actual demand, but also has high efficiency.

[0066] In addition, the second constraint condition is another constraint on the neural network during the training process, which is used to encourage the model to generate a path that arrives at the network point on time or in advance. Specifically, for a path that arrives at the network point in advance, the model will obtain a positive reward; and for a path that arrives at the network point overtime, the model will receive a negative reward. This reward mechanism encourages the model to learn how to complete the task within the specified time.

[0067] As can be seen from the above, in the embodiments of the present application, the neural network is iteratively trained according to the training samples, the first constraint condition and the second constraint condition. The first constraint condition ensures that the path generated by the model not only meets the actual demand, such as satisfying the network point priority, but also has high efficiency, i.e., the shortest time consumption. The second constraint condition encourages the model to learn how to complete the task within the specified time through the reward mechanism, thereby improving the punctuality and reliability of the path planning. Through the positive reward and negative reward, the model can better learn how to make the optimal decision in a complex environment. This reinforcement learning mechanism enables the model to dynamically adapt to different task requirements. The target model obtained through training can dynamically adjust the path planning according to the input real-time data, such as traffic conditions and weather conditions, and has strong adaptability.

[0068] Optionally, the target feature matrix is constructed according to the route planning reference information, the route graph, the set priority of each network point for the cash transport task, and the vehicle state data of the cash transport vehicle, including: the cash transport vehicle navigation system can create an initial feature matrix, wherein the rows of the initial feature matrix represent each road node in the route graph and N network points, and the columns represent the feature information to be filled in. Then, the cash transport vehicle navigation system can fill in the corresponding data for each cell of the initial feature matrix to obtain the target feature matrix, wherein for each road node in the route graph, the average speed, the congestion index and the weather influence coefficient when the cash transport vehicle travels to each road node are filled in; and for each network point, the set priority of the network point, the predicted cash transport vehicle arrival time and the probability of occurrence of a public security event are filled in.

[0069] Optionally, the initial feature matrix is the initial state of the target feature matrix, and the structure thereof has been defined, but specific data has not been filled in. The rows of the initial feature matrix correspond to each road node and network point in the route graph, and the columns correspond to the feature information to be filled in.

[0070] Optionally, the feature information refers to various attribute data related to path planning, such as average speed of road nodes, congestion index, weather influence coefficient, and priority of network nodes, predicted arrival time, and probability of public security events, etc. These feature information is used to describe the characteristics of roads and network nodes, helping the model to plan the path. The average speed can be obtained by historical data or real-time traffic data, which is used to evaluate the traffic efficiency of the road. The congestion index represents the congestion degree of the road section. The higher the congestion index, the more congested the road, and the slower the vehicle speed. The weather influence coefficient refers to the influence coefficient of weather on road traffic. For example, in bad weather, the traffic capacity of the road will decrease, and the influence coefficient will increase.

[0071] From the above, in the embodiments of the present application, by creating an initial feature matrix and filling the corresponding data for each cell of the initial feature matrix, a target feature matrix is obtained, which can integrate the route map, network node priority, vehicle state data and route planning reference information together by creating the target feature matrix, and provide comprehensive data support for path planning.

[0072] In addition, the target feature matrix not only contains the traffic efficiency of the road, such as average speed and congestion index, but also considers weather influence, network node priority and public security risk, etc. This fine-grained feature extraction makes the path planning more scientific and comprehensive. The data in the target feature matrix can be dynamically updated according to real-time traffic, weather and public security conditions, so that the path planning can adapt to complex environmental changes in real time. Through the target feature matrix, the path planning model can consider multiple factors and generate the optimal driving path. For example, the model can preferentially select the path with low congestion index and small public security risk, while satisfying the constraints of network node priority and vehicle state.

[0073] According to another aspect of the embodiments of the present application, a path determination device is also provided, Figure 3 is a schematic diagram of an optional path determination device according to an embodiment of the present application, as Figure 3 shown, the device comprises a first determination unit 301, a first acquisition unit 302 and a second determination unit 302.

[0074] The first determining unit 301 is configured to determine a route map according to the location of the cash transport vehicle and the locations of the N banking outlets, wherein N is an integer greater than 1, and the route map at least includes path information and traffic information from the cash transport vehicle to each of the N banking outlets and path information and traffic information between any two of the N banking outlets; the first obtaining unit 302 is configured to obtain a set priority of each of the N banking outlets for the cash transport task; and the second determining unit 303 is configured to determine a target driving path of the cash transport vehicle according to the route map, the set priority of each of the N banking outlets for the cash transport task, and vehicle state data of the cash transport vehicle, wherein the target driving path includes the N banking outlets.

[0075] Optionally, the first determining unit 301 includes a first obtaining subunit configured to obtain geographic location information of a target location, wherein the target location is a location with a crowd density greater than a preset threshold and / or a location with a frequency of public security events greater than a preset frequency; and a first processing subunit configured to mark the geographic location information of the target location in the route map and establish a path connection relationship between the target location and the location of the cash transport vehicle and the locations of the N banking outlets.

[0076] Optionally, the second determining unit 302 includes a collecting subunit configured to collect route planning reference information, wherein the route planning reference information at least includes meteorological data, a historical driving route of the cash transport vehicle in performing a historical cash transport task, and time consumption information based on the historical driving route to reach each of the N banking outlets; and a first determining subunit configured to determine the target driving path of the cash transport vehicle according to the route planning reference information, the route map, the set priority of each of the N banking outlets for the cash transport task, and the vehicle state data of the cash transport vehicle.

[0077] Optionally, the first determining subunit includes a constructing module configured to construct a target feature matrix according to the route planning reference information, the route map, the set priority of each of the N banking outlets for the cash transport task, and the vehicle state data of the cash transport vehicle; and a first determining module configured to input the target feature matrix into a target model to determine the target driving path of the cash transport vehicle by prior knowledge learned by the target model in a training process.

[0078] Optionally, the first determining module includes a first determining sub-module configured to determine a driving environment of the cash transport vehicle by the target model according to the target feature matrix; a first processing sub-module configured to take a driving action of the cash transport vehicle as an operation behavior in the driving environment; and a second determining sub-module configured to determine the target driving path of the cash transport vehicle by the prior knowledge of the target model according to the driving environment and the operation behavior, wherein the target model learns a target operation of the cash transport vehicle in different driving environment states in the training process, and the target operation ensures that the cash transport vehicle drives through the N banking outlets in the shortest time.

[0079] Optionally, the path determination apparatus further comprises: a sample obtaining unit, configured to obtain training samples, wherein the training samples comprise M reference routes formulated based on the route graph, M being an integer greater than 1; and a training unit, configured to perform multiple iterative training on the neural network according to the training samples, the first constraint condition and the second constraint condition, until the neural network enters a convergent state, and obtain the target model by training, wherein the first constraint condition is used to constrain the neural network to design a new route according to the M reference routes, and the design target is to design a route with the shortest time consumption under the condition that the set priority of each network point for the cash transport task is met; and the second constraint condition is used to set a positive reward for reaching a network point in advance and set a negative reward for reaching a network point overtime in the process of designing the route.

[0080] Optionally, the construction module comprises: a creation submodule, configured to create an initial feature matrix, wherein the rows of the initial feature matrix represent each road node in the route graph and the N network points, and the columns represent the feature information to be filled; and a filling submodule, configured to fill each cell of the initial feature matrix with corresponding data to obtain the target feature matrix, wherein for each road node in the route graph, the average speed, the congestion index and the weather influence coefficient when the cash transport vehicle drives to each road node are filled; and for each network point, the set priority, the predicted cash transport vehicle arrival time and the probability of occurrence of a public security event of the network point are filled.

[0081] According to another aspect of the embodiments of the present application, a computer readable storage medium is further provided, wherein the computer readable storage medium stores a computer program, and when the computer program runs, the computer readable storage medium causes the device where the computer readable storage medium is located to perform the path determination method.

[0082] According to another aspect of the embodiments of the present application, an electronic device is further provided, wherein the electronic device comprises one or more processors and a memory, and the memory is configured to store one or more programs, and when the one or more programs are executed by the one or more processors, the one or more processors perform the path determination method.

[0083] According to another aspect of the embodiments of the present application, a computer program product is further provided, wherein the computer program product comprises a computer program or instructions, and when the computer program or instructions are executed by a processor, the path determination method is implemented.

[0084] The above sequence numbers of the embodiments of the present application are only for description, and do not represent the advantages and disadvantages of the embodiments.

[0085] In the above embodiments of the present application, the description of each embodiment has its own focus, and the parts not described in detail in a certain embodiment can be referred to the related description of other embodiments.

[0086] In several embodiments provided in the present application, it should be understood that the disclosed technology can be implemented by other ways. Among them, the above-described device embodiments are only schematic, for example, the division of the units can be a logical function division, and actual implementation can have another division manner, for example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the displayed or discussed units can be indirect coupling or communication connection through some interfaces, units or modules, and can be electrical or other forms.

[0087] The units described as separate components can or can not be physically separate, and the components shown as units can or can not be physical units, that is, they can be located in one place, or they can be distributed to multiple units. Part or all of the units can be selected according to actual needs to achieve the purpose of the embodiment.

[0088] In addition, the functional units in each embodiment of the present application can be integrated in one processing unit, or each unit can be physically present separately, or two or more units can be integrated in one unit. The integrated unit can be realized in the form of hardware or in the form of a software functional unit.

[0089] The integrated unit, if realized in the form of a software functional unit and sold or used as an independent product, can be stored in a computer readable storage medium. Based on this understanding, the technical solutions of the present application essentially or the part of the prior art that contributes to the technical solutions or the whole or part of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server or a network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The foregoing storage medium includes: a U disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a mobile hard disk, a magnetic disk or an optical disk, and various media that can store program codes.

[0090] The above is only the preferred embodiment of the present application, and it should be pointed out that for ordinary skilled in the art, without departing from the principles of the present application, a number of improvements and refinements can be made, and these improvements and refinements should be considered as the protection scope of the present application.

Claims

1. A path determination method, characterized in that: include: Determining a route map based on the location of the armored car and the locations of N outlets, wherein N is an integer greater than 1, and the route map includes at least: path information and traffic information from the armored car to each outlet, and path information and traffic information between any two outlets among the N outlets; Obtaining the set priority of each branch for the cash transport task; The target driving path of the cash transport vehicle is determined according to the route map, the set priority of each network point for the cash transport task, and the vehicle status data of the cash transport vehicle, wherein the target driving path includes the N network points.

2. The path determination method according to claim 1, wherein: Based on the location of the armored car and the locations of N outlets, a route map is determined, including: Obtaining geographic location information of a target location, wherein the target location is a location where a crowd density is greater than a preset threshold and / or a location where a frequency of public security incidents occurs is greater than a preset frequency; The geographical location information of the target location is marked on the route map, and a path connection relationship is established between the target location and the location of the armored car and the location of each network point.

3. The path determination method according to claim 1, wherein: Determining a target driving path of the cash transport vehicle according to a route map, the priority set by each network point for the cash transport task, and the vehicle status data of the cash transport vehicle includes: Collecting route planning reference information, wherein the route planning reference information includes at least: weather data, historical driving routes of the cash transport vehicle when performing historical cash transport tasks, and time information for reaching each network point based on the historical driving routes; The target driving path of the cash transport vehicle is determined according to the route planning reference information, the route map, the set priority of each network point for the cash transport task, and the vehicle status data of the cash transport vehicle.

4. The path determination method according to claim 3, wherein: Determining a target driving path of the cash transport vehicle according to the route planning reference information, the route map, the priority set by each network point for the cash transport task, and the vehicle status data of the cash transport vehicle includes: Constructing a target feature matrix based on the route planning reference information, the route map, the set priority of each network point for the cash transport task, and the vehicle status data of the cash transport vehicle; The target feature matrix is ​​input into the target model, and the target driving path of the armored car is determined by the prior knowledge learned by the target model during the training process.

5. The path determination method according to claim 4, characterized in that: The target feature matrix is ​​input into the target model, and the target driving path of the cash transport vehicle is determined based on the prior knowledge learned by the target model during the training process, including: Determining the driving environment of the money transport vehicle according to the target feature matrix through a target model; The driving action of the cash transport vehicle is regarded as the operation behavior in the driving environment; The target driving path of the armored car is determined based on the driving environment and the operating behavior through the prior knowledge of the target model, wherein the target model learns the target operation of the armored car under different driving environment conditions during the training process, wherein the target operation ensures that the armored car passes through the N outlets in the shortest time.

6. The path determination method according to claim 5, characterized in that: The training process of the target model includes the following steps: Acquire a training sample, wherein the training sample includes M reference routes developed based on the route map, where M is an integer greater than 1; Based on the training samples, the first constraint and the second constraint, the neural network is iteratively trained multiple times until the neural network enters a convergence state, and the target model is obtained by training, wherein the first constraint is used to constrain the neural network to design a new route based on the M reference routes, and the design goal is to design a route with the shortest time while satisfying the set priority of each branch for the cash transportation task; the second constraint is used to set a positive reward for arriving at the branch ahead of time and a negative reward for arriving at the branch after time in the process of designing the route.

7. The path determination method according to claim 4, characterized in that: A target feature matrix is ​​constructed based on the route planning reference information, the route map, the priority set by each network point for the cash transport task, and the vehicle status data of the cash transport vehicle, including: Creating an initial feature matrix, wherein the rows of the initial feature matrix represent the various road locations and the N network points in the route map, and the columns represent the feature information to be filled; The target feature matrix is ​​obtained by filling corresponding data into each cell of the initial feature matrix, wherein, for each road node in the route map, the average speed, congestion index and weather impact coefficient of the armored car when traveling to each road node are filled in; and for each network point, the set priority of the network point, the predicted arrival time of the armored car and the probability of a public security incident are filled in.

8. A path determination device, characterized in that: include: a first determining unit configured to determine a route map based on the location of the armored car and the locations of N outlets, wherein N is an integer greater than 1, and the route map includes at least: path information and traffic information from the armored car to each outlet, and path information and traffic information between any two outlets among the N outlets; A first acquiring unit is configured to acquire a set priority of each branch for the cash transport task; The second determining unit is used to determine the target driving path of the cash transport vehicle according to the route map, the set priority of each network point for the cash transport task, and the vehicle status data of the cash transport vehicle, wherein the target driving path includes the N network points.

9. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, wherein when the computer program is executed, the device where the computer-readable storage medium is located executes the path determination method according to any one of claims 1 to 7.

10. An electronic device, characterized in that: The method comprises one or more processors and a memory, wherein the memory is used to store one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors execute the path determination method according to any one of claims 1 to 7.

11. A computer program product, characterized in that The method comprises a computer program or an instruction, which implements the path determination method according to any one of claims 1 to 7 when executed by a processor.

Images