A path planning method and system for a road maintenance vehicle

Abstract

The present invention discloses a path planning method and system for a road maintenance vehicle, which relates to the field of artificial intelligence technology. It includes collecting multi-modal data of the road maintenance vehicle through sensors and 5G network and uploading it to the cloud data center for multi-modal data fusion; based on the fused multi-modal data, using the quantum optimization algorithm to quickly calculate the optimal path set; performing maintenance tasks; for the road maintenance vehicle performing tasks on the road, the cloud data center will continuously receive feedback data; after completing the tasks, the cloud data center will summarize the operation data of the road maintenance vehicle and conduct in-depth analysis. The present invention integrates multi-modal sensors and 5G technology, monitors the environment in real time, uses the quantum optimization and CADPOA algorithms to calculate and dynamically adjust the optimal path, automatically performs tasks, and the feedback mechanism continuously optimizes the path planning, realizing continuous and efficient maintenance work, and ensuring the continuity and high quality of the maintenance work.

Description

Technical Field

[0001] The present invention relates to the technical field of artificial intelligence, and particularly to a path planning method and system for a road maintenance vehicle. Background Art

[0002] The intelligent transformation of road maintenance is a core issue in modern traffic management. In recent years, with the rapid development of Internet of Things (IoT) technology and 5G communication, multi-modal data collection and analysis have become possible, providing a data foundation for the efficient operation of road maintenance vehicles. The low latency and high bandwidth characteristics of the 5G network ensure the reliability of real-time data transmission. The application of multi-modal sensors, such as high-definition cameras, LiDAR, GPS, IMU, and various environmental monitoring sensors, comprehensively captures diverse information such as road conditions, traffic flow, and vehicle status, laying a solid foundation for refined management. The progress of data fusion technology enables the effective integration of data from different sensors to form a unified situation awareness, further promoting the improvement of intelligent decision-making systems.

[0003] However, despite the significant progress made in the existing technology at the data collection and transmission level, the limitations of path planning algorithms still restrict the overall improvement of road maintenance efficiency. Traditional path planning algorithms, such as Dijkstra and AI algorithms, although performing well in static environments, are difficult to cope with complex dynamic road conditions and changing maintenance requirements. They often ignore the multi-modal characteristics of data, resulting in inaccurate path planning and the inability to fully utilize real-time information to adjust the optimal path. In addition, the lack of an efficient real-time path adjustment mechanism causes the maintenance vehicle to react slowly in the face of emergencies, affecting the continuity and safety of operations. Therefore, there is an urgent need for a new path planning method that can quickly process multi-modal data, has real-time optimization capabilities, and high adaptability to significantly improve the intelligent level and operation efficiency of road maintenance. Summary of the Invention

[0004] In view of the above existing problems, the present invention is proposed.

[0005] Therefore, the present invention provides a path planning method and system for a road maintenance vehicle to solve the problems of data fusion, dynamic optimization, and intelligent decision-making in the path planning of road maintenance vehicles.

[0006] To solve the above technical problems, the present invention provides the following technical solutions:

[0007] In a first aspect, an embodiment of the present invention provides a path planning method for a road maintenance vehicle, which includes collecting multi-modal data of the road maintenance vehicle through sensors and a 5G network and uploading it to a cloud data center for multi-modal data fusion; based on the fused multi-modal data, using a quantum optimization algorithm to quickly calculate an optimal path set; on the basis of calculating the optimal path, using the CADPOA algorithm for real-time path prediction and adjustment and uploading it to the cloud data center through the 5G network; the road maintenance vehicle plans its driving route by itself according to the optimal path instruction issued by the cloud data center and executes maintenance tasks; for the road maintenance vehicle performing tasks on the road, the cloud data center will continuously receive feedback data; after the task is completed, the cloud data center will summarize the operation data of the road maintenance vehicle and conduct in-depth analysis.

[0008] As a preferred solution of the path planning method for the road maintenance vehicle described in the present invention, wherein: the step of collecting multi-modal data of the road maintenance vehicle through sensors and a 5G network and uploading it to the cloud data center for multi-modal data fusion is specifically as follows,

[0009] The road maintenance vehicle collects multi-modal data in real time through multi-modal sensors;

[0010] Preprocess the multi-modal data;

[0011] Upload the preprocessed multi-modal data to the cloud data center in real time through the 5G network;

[0012] Use the EBFFA algorithm to fuse the preprocessed data set.

[0013] As a preferred solution of the path planning method for the road maintenance vehicle described in the present invention, wherein: the step of, based on the fused multi-modal data, using a quantum optimization algorithm to quickly calculate an optimal path set is specifically as follows,

[0014] Based on the fused multi-modal data, use the QEMOOA algorithm to calculate the optimal path at time t.

[0015] As a preferred solution of the path planning method for the road maintenance vehicle described in the present invention, wherein: the step of, on the basis of calculating the optimal path, using the CADPOA algorithm for real-time path prediction and adjustment and uploading it to the cloud data center through the 5G network is specifically as follows,

[0016] Based on the optimal path, use the CADPOA algorithm for real-time path adjustment and prediction;

[0017] Upload the adjusted and predicted real-time path to the cloud data center through the 5G network.

[0018] As a preferred solution of the path planning method for the road maintenance vehicle described in the present invention, wherein: the road maintenance vehicle plans its own driving route according to the optimal path instruction issued by the cloud data center and executes the maintenance task. The specific steps are as follows:

[0019] After receiving the optimal path instruction issued by the cloud data center, the road maintenance vehicle first parses the instruction, including the details of the maintenance tasks to be executed by the maintenance vehicle, the expected driving route, the maintenance demand points along the way, the expected arrival time, as well as safety tips and warning information;

[0020] Based on the parsed instruction, the on-vehicle navigation of the road maintenance vehicle will automatically plan the driving route;

[0021] And generate a detailed driving plan, including the departure time, the expected arrival time, turning points, deceleration areas, acceleration areas, and the staying time at the maintenance points;

[0022] The maintenance vehicle executes the maintenance task according to the planned route, and the on-vehicle software monitors the real-time state of the vehicle.

[0023] As a preferred solution of the path planning method for the road maintenance vehicle described in the present invention, wherein: for the road maintenance vehicle performing tasks on the road, the cloud data center will continuously receive feedback data. The specific steps are as follows:

[0024] The cloud data center analyzes in real time the data uploaded by the road maintenance vehicle performing tasks on the road;

[0025] Dynamically adjust the path planning and evaluate its performance, including travel time, fuel efficiency, safety index, path deviation, task completion rate;

[0026] When the actual travel time exceeds 10% of the expected time, trigger path replanning;

[0027] When the fuel efficiency is lower than 12 kilometers per liter of gasoline, adjust the speed;

[0028] When the safety index deteriorates, preventive measures should be taken.

[0029] As a preferred solution of the path planning method for the road maintenance vehicle described in the present invention, wherein: after the task is completed, the cloud data center will summarize the operation data of the road maintenance vehicle and conduct in-depth analysis. The specific steps are as follows:

[0030] After the road maintenance vehicle completes the task, the cloud data center summarizes the operation data of all road maintenance vehicles, including vehicle status, operation efficiency, environmental factors, road conditions information, operation path, safety index, task completion rate;

[0031] Preprocess the summarized data;

[0032] Construct an analysis framework and conduct in-depth analysis for different performance indicators;

[0033] Use time series models to identify the efficiency in different time periods and test the average efficiency between different job types through variance analysis;

[0034] Calculate the benefits brought by unit cost and compare the cost-benefits of different job types and vehicles;

[0035] Use the geographic information GIS tool to compare the geographical coordinates of the actual and planned routes, and analyze traffic conditions and weather changes by cross-referencing real-time traffic data;

[0036] Utilize decision tree models to predict the best route under future road conditions and use moving averages to identify trends in safety metrics over time;

[0037] Identify abnormal safety events through statistical tests and finally use the Kaplan-Meier survival curve to estimate the duration of maintenance effects.

[0038] In a second aspect, the present invention provides a path planning system for a road maintenance vehicle, including a data collection module, a fusion module, a path calculation module, a prediction and adjustment module, an execution feedback module, and an analysis and optimization module; the data collection module is used to collect multi-modal data of the road maintenance vehicle through sensors and a 5G network and upload it to the cloud data center for multi-modal data fusion; the fusion module is used to quickly calculate an optimal path set based on the fused multi-modal data using a quantum optimization algorithm; the path calculation module is used to perform real-time path prediction and adjustment using the CADPOA algorithm based on the calculated optimal path and upload it to the cloud data center using a 5G network; the prediction and adjustment module is used for the road maintenance vehicle to independently plan a driving route and perform maintenance tasks according to the optimal path instruction issued by the cloud data center; the execution feedback module is used for the cloud data center to continuously receive feedback data for the road maintenance vehicle performing tasks on the road; the analysis and optimization module is used to summarize the operation data of the road maintenance vehicle and conduct in-depth analysis after the task is completed.

[0039] In a third aspect, an embodiment of the present invention provides a computer device, including a memory and a processor, where the memory stores a computer program, and: when the computer program is executed by the processor, any step of the path planning method for a road maintenance vehicle as described in the first aspect of the present invention is implemented.

[0040] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium, on which a computer program is stored, and: when the computer program is executed by the processor, any step of the path planning method for a road maintenance vehicle as described in the first aspect of the present invention is implemented.

[0041] The beneficial effects of the present invention are as follows: By integrating multi-modal sensors and 5G networks, the present invention realizes all-round real-time monitoring of the environment around the maintenance vehicle, provides detailed data for path planning, and uses the quantum optimization algorithm. This method can quickly calculate the optimal path set under multi-objective constraints, significantly reduce the driving time and fuel consumption of maintenance operations, and improve safety at the same time. With the help of the CADPOA algorithm, the path planning has the ability of real-time adjustment and prediction, ensuring that the maintenance vehicle can still maintain efficient operation in the face of emergencies. The maintenance vehicle automatically executes maintenance tasks according to the optimal path instructions issued by the cloud data center, improving the standardization and safety of operations. The entire system dynamically adjusts the path planning by continuously receiving feedback data, ensuring the continuity and high quality of maintenance work. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0043] Figure 1 It is a flowchart of the path planning method for the road maintenance vehicle in Embodiment 1.

[0044] Figure 2 It is a flowchart of the automated operation in Embodiment 1. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0045] In order to make the above objects, features and advantages of the present invention more obvious and understandable, the following will make a detailed description of the specific embodiments of the present invention with reference to the drawings in the specification.

[0046] Embodiment 1, referring to Figure 1 and Figure 2 , is the first embodiment of the present invention. This embodiment provides a path planning method for a road maintenance vehicle, including the following steps:

[0047] S1. Collect multi-modal data of the road maintenance vehicle through sensors and 5G networks and upload them to the cloud data center for multi-modal data fusion.

[0048] Furthermore, the road maintenance vehicle collects multi-modal data in real time through multi-modal sensors;

[0049] The multi-modal data includes:

[0050] High-definition camera data: Provide visual information for identifying road conditions (such as cracks, potholes), obstacles (such as falling rocks, animals), signs and markings, surrounding environment (such as weather, lighting conditions), etc.;

[0051] LiDAR data: Measure distance and position information for constructing a high-precision 3D map, detecting static and dynamic obstacles, and evaluating road surface flatness;

[0052] Ultrasonic sensor data: Used for close-range obstacle detection, especially under low visibility conditions, providing additional safety;

[0053] Thermal imaging camera data: Can detect heat sources such as engine overheating, heat generated by tire wear, and possible organisms at night or under low light conditions;

[0054] GPS / GNSS data: Provide accurate position information for vehicle positioning and recording driving trajectories;

[0055] Inertial measurement unit (IMU) data: Include accelerometers and gyroscopes for measuring vehicle acceleration, tilt angle, and rotation, assisting in positioning and attitude estimation;

[0056] Traffic flow monitoring data: Evaluate traffic density and predict potential congestion by detecting the number, speed, and type of passing vehicles;

[0057] Meteorological sensor data: Include temperature, humidity, wind speed, and rainfall, etc., for evaluating the impact of weather conditions on operations;

[0058] Vehicle status data: Include engine speed, oil pressure, coolant temperature, fuel level, battery voltage, etc., for monitoring vehicle health;

[0059] Communication data: Include communication information with other vehicles or infrastructure for cooperative operations and collision avoidance;

[0060] Preprocess multi-modal data;

[0061] The data is initially cleaned and format-converted to remove noise and redundancy, ensuring data quality. Specific preprocessing includes:

[0062] Remove duplicates, fill in missing values, correct incorrect data, and handle outliers to improve data quality;

[0063] Encode and convert non-numerical data into numerical data to improve data distribution;

[0064] Scale the data to the same scale through min-max normalization or Z-score normalization to eliminate the influence of dimension;

[0065] Identify the feature subset most relevant to the target, remove irrelevant or redundant features to reduce model complexity and improve model performance. Feature selection methods include:

[0066] Filter methods: Use chi-square test, variance threshold, and correlation coefficient to evaluate the correlation between features and the target variable;

[0067] Wrapper methods: Evaluate the performance of different feature subsets on a specific model through recursive feature elimination (RFE) and genetic algorithms;

[0068] Embedded methods: Perform feature selection during model training through regularization techniques such as LASSO and Ridge regression, and decision tree algorithms such as random forest and GBDT;

[0069] Data filling strategies should be considered to address possible sensor failures or data loss;

[0070] Preferably, common encoding methods include:

[0071] One-hot encoding: For categorical variables, convert them into binary vectors, where only one bit represents 1 and the rest are 0, indicating the category of the categorical variable;

[0072] Label encoding: Convert categorical variables into integers, with each category corresponding to an integer, but this method may introduce unnecessary order relationships;

[0073] Ordinal encoding: Similar to label encoding, but applicable to ordered categorical variables, where the magnitude of the numbers reflects the order of the categories;

[0074] Embedding encoding: Use the embedding layer of a neural network to map categorical variables into a continuous vector space, and this encoding method can capture the potential relationships between categories;

[0075] Upload the preprocessed multi-modal data to the cloud data center in real time through the 5G network;

[0076] Adopt the EBFFA algorithm to fuse the preprocessed data set, and the expression is:

[0077]

[0078] Among them, F(D(t)) is the output value of the fusion algorithm, representing the result of fusing all data D at time t, and H i (D i (t)) is the information entropy value of the i-th sensor data D i (t), which is used to measure the uncertainty and information content of the data, and F i (D i (t)) is the i-th sensor data D i(t) is the fuzzy membership function value, which reflects the correlation and credibility between the data and the fusion target. i is the index of the sensor data, and n is the total number of sensor data.

[0079] Preferably, H i (D i (t)) is the information entropy value of the i-th sensor data D i (t). Information entropy is a statistical measure that quantifies the uncertainty or amount of information in data. For a discrete random variable X, the information entropy is defined as H(X), and the expression is:

[0080] H(X) = -∑ x∈X P(x) log 2 P(x);

[0081] Among them, H(X) represents the information entropy of the random variable X, which is also a measure of the uncertainty of the random variable X. The higher the information entropy, the greater the uncertainty of the random variable, that is, the more information it contains. ∑ x∈X represents the sum over all possible values of x. Here, x is a value taken by the random variable X. For example, if X is the result of rolling a die, then the range of x is {1, 2, 3, 4, 5, 6}. P(x) represents the probability that the random variable X takes the value x. This probability must satisfy 0 ≤ P(x) ≤ 1, and the sum of the probabilities for all x ∈ X is equal to 1. For example, if X represents the result of flipping a fair coin, then P(heads) = P(tails) = 0.5. log 2 P(x) represents the logarithm to the base 2. Here, 2 is used because information is usually measured in bits. The role of the logarithmic function here is to convert the probability into a measure of the amount of information. Because when the probability P(x) is close to 1, that is, the event is almost certain to occur, the amount of information is small; while when P(x) is close to 0, that is, the event is almost certain not to occur, the amount of information is large. Using the logarithm to the base 2 ensures that the unit of the amount of information is bits. The negative sign in the information entropy formula is because the range of the logarithmic function is from negative infinity to zero. Adding the negative sign makes the range of the information entropy from zero to positive infinity, so that the information entropy becomes a positive value, representing the amount of information.

[0082] For the sensor data D i (t), first count the frequency of each possible state and its occurrence, and then substitute it into the above formula to calculate the information entropy. If the data is continuous, it usually needs to be discretized first;

[0083] Preferably, F i (D i (t)) is the i-th sensor data D iThe fuzzy membership function value of (t), where the fuzzy membership function is used to evaluate the relevance and credibility of data with respect to the fusion target;

[0084] The calculation of the fuzzy membership function depends on the specific application scenario and data type. Generally, it is calculated based on the similarity between the data and known standards or patterns, such as through distance metrics or similarity metrics. For a specific sensor data, the membership degree can be determined by comparing the difference between its reading and the ideal or typical reading;

[0085] It should be noted that by collecting and uploading the environmental and vehicle status data of the road maintenance vehicle in real time through multi-modal sensors to the cloud, after preprocessing and fusion, a unified data view is formed. This process improves the data quality, ensures the accuracy of decision-making, supports efficient road condition analysis and vehicle health management, realizes intelligent path planning and preventive maintenance, and ultimately optimizes the efficiency and safety of maintenance operations.

[0086] S2. Based on the fused multi-modal data, use the quantum optimization algorithm to quickly calculate the optimal path set.

[0087] Furthermore, based on the fused multi-modal data, use the QEMOOA algorithm to calculate the optimal path at time t, and the expression is:

[0088]

[0089] The objective function, and the expression is:

[0090] G = 3;

[0091] Among them, P * is the optimal path selection value, is the path value P that finds the maximum value of the expression in the parentheses, and Q g (P, t) is the quantum expected value of the g-th objective function at path P and time t, reflecting the optimization ability of quantum computing in the probability space. γ g is the quantum weight factor of the g-th objective function, reflecting the importance of the objective function. g is the index of the objective function, ranging from 1 to G, and G is the total number of objective functions. 3 is three special values in the objective function, corresponding to the total travel time value T(P), the total fuel consumption value E(P), and the total safety value S(P) respectively.

[0092] Preferably, the objective function is defined as follows: Establish multiple objective functions covering key indicators such as minimizing travel time, minimizing fuel consumption, and maximizing safety. These objective functions can be expressed in mathematical form as:

[0093] Minimizing travel time: minT(P) = ∑ e∈P t e ;

[0094] Minimize fuel consumption: minE(P) = ∑ e∈P e e ;

[0095] Maximize safety: maxS(P) = ∑ e∈P s e ;

[0096] Where T(P) represents the total travel time on path P, which is one of the objective functions to be minimized in path planning with the aim of finding the path with the least time consumption. e refers to an edge or section in path P. In the context of a road maintenance vehicle, each edge can represent the section between two inspection points or locations. P represents a path value composed of a series of connected edges, representing the driving route of the road maintenance vehicle from the starting point to the ending point. t e is the travel time of edge e, that is, the time required for the road maintenance vehicle to pass through section e. This time can be affected by various factors such as road conditions, traffic flow, speed limits, etc. E(P) represents the total fuel consumption on path P, which is another objective function to be minimized with the aim of finding the path with the least fuel consumption. e e is the energy consumption or fuel consumption of edge e. This value may depend on factors such as the length of the section, road conditions (such as slope, road surface type), vehicle speed, and load. S(P) represents the overall safety value of path P, which is an objective function to be maximized with the aim of finding the safest path. s e is the safety score of edge e, which can be evaluated based on various factors such as accident rate, sight conditions, road maintenance status, traffic density, etc. The ∑ symbol represents the summation over all edges in path P.

[0097] Preferably, γ g This parameter is usually called the quantum weight factor, which reflects the importance of each objective function; in a multi-objective optimization problem, each objective may have different priorities, and γ g is used to quantify this priority. In practice, the value of γ g can be determined by expert knowledge, empirical rules, or through sensitivity analysis. It can also be an adjustable parameter, allowing users to adjust the relative importance of the optimization objectives according to the actual situation;

[0098] γ g can be set based on the opinions of domain experts, can be adjusted according to solutions to previous similar problems, can observe the changes in results by changing the value of γ g to find the most sensitive parameter, and can take γ g as an additional optimization variable and adjust it through, for example, genetic algorithms to obtain an optimal or approximate optimal solution;

[0099] Qg (P, t) This expression refers to the quantum expectation value of the g-th objective function under path P and time t. The concept of quantum expectation value originates from the expectation value theory in quantum mechanics and is used to evaluate the average performance of the objective function in the probability space. In multi-objective optimization, this means that for each possible path and a given time point, we can calculate the expected contribution of that path to the objective function. The calculation of this value usually involves the principles of quantum computing, such as quantum state superposition and quantum entanglement, to simulate and optimize the possible solution space;

[0100] Q g (P, t) is obtained by measuring the expectation value of the wave function. In a quantum circuit, the quantum expectation value of the objective function is usually realized by constructing a quantum circuit that can prepare a quantum state related to the objective function. Then, the expectation value of this quantum state is estimated through quantum measurement, which usually involves quantum gate operations such as Hadamard gates, controlled gates, and rotation gates, as well as the phenomena of quantum state superposition and entanglement;

[0101] Preferably, the quantum optimization algorithm utilizes the characteristics of quantum computing, such as superposition and entanglement, to search for the optimal solution. In quantum computing, qubits can be in multiple states simultaneously, which enables quantum computers to process multiple computational paths simultaneously and thus theoretically find the global optimal solution faster than classical computers;

[0102] QEMOOA is a quantum-inspired multi-objective optimization algorithm that combines the principles of quantum computing and the ideas of evolutionary algorithms. It explores the solution space through the superposition and entanglement of qubits and uses quantum gate operations (such as Hadamard gates) to update the solution set to find a set of Pareto optimal solutions;

[0103] It should be noted that based on the fused multi-modal data, the QEMOOA algorithm utilizes the advantages of quantum computing, adjusts the objective priorities through quantum weight factors, evaluates the multi-objective optimization of paths with quantum expectation values, and quickly finds the optimal path set. This method combines quantum superposition and entanglement, breaks through the limitations of traditional computing, and greatly improves the efficiency and quality of path planning. It is especially suitable for complex scenarios that need to consider multi-dimensional objectives such as time, cost, and safety, and realizes the intelligence and optimization of highway maintenance operations.

[0104] S3. On the basis of calculating the optimal path, use the CADPOA algorithm for real-time path prediction and adjustment and upload it to the cloud data center using the 5G network.

[0105] Furthermore, based on the optimal path, use the CADPOA algorithm for real-time path adjustment and prediction to cope with emergencies or unforeseen changes. The expression is:

[0106]

[0107] Among them, P n (t) is the new adjusted n - path value at time t, is the specific path value that finds the minimum value among all possible paths, indicating finding the path P(t) that makes the expression in the brackets minimum among all possible paths P(t), α t is the weight factor of the travel time varying with time t, β t is the weight factor of the fuel consumption varying with time t, γ t is the weight factor of the maintenance efficiency varying with time t, T(P, t) represents the total travel time of path P at time t, which is usually determined by the speed of the maintenance vehicle, the path length, and the possible traffic conditions, E(P, t) represents the total fuel consumption value of path P at time t, which may depend on the fuel efficiency of the vehicle, the load, and the road conditions, C(P, t) represents the maintenance efficiency value of path P at time t, that is, the speed and quality of completing the maintenance task, which may be affected by the type of the maintenance vehicle, the nature of the maintenance task, and the suitability of the maintenance vehicle for a specific task, δ is a constant weight factor used to balance the certainty of the path and the adaptability to unknown events, V(P, t) represents the uncertainty value of path P at time t, which takes into account the possibility of future road condition changes, such as traffic jams, weather changes, or sudden maintenance requirements, and the uncertainty V can be a probability distribution predicted based on historical data and real - time information;

[0108] Upload the adjusted and predicted real - time path to the cloud data center through the 5G network;

[0109] Preferably, α t represents the weight factor of the travel time varying with time t, and its value can be dynamically adjusted according to the current task urgency of the maintenance vehicle, the change of traffic conditions, or the schedule of the maintenance vehicle. For example, if the maintenance vehicle needs to complete the task as soon as possible, then α t may increase in value to prioritize reducing the travel time. During the traffic peak period, α t may be adjusted higher to avoid extending the travel time due to traffic congestion;

[0110] β t is the weight factor of the fuel consumption varying with time t, reflecting the importance of fuel consumption at time t, and it may change with the fluctuation of fuel prices, the remaining fuel quantity of the vehicle, or environmental policy requirements. For example, when the fuel cost rises, β t may be set higher to encourage the maintenance vehicle to choose a more energy - efficient path. During a long - distance task, when the fuel reserve is low, β t may also be adjusted higher to avoid the need for refueling on the way;

[0111] γt is a weight factor for maintaining efficiency that varies with time t. This parameter can be adjusted according to the maintenance task type of the maintenance vehicle, the priority of the task, and the maintenance ability of the maintenance vehicle itself. For example, when the maintenance vehicle performs a maintenance task with a higher criticality, γ t may be set relatively high to ensure the efficient completion of the task. On the other hand, if the maintenance resources of the maintenance vehicle are limited, then γ t may be reduced to avoid excessive consumption of resources;

[0112] δ is a constant weight factor used to balance the certainty of the path and the adaptability to unknown events. This parameter usually does not change with time, but it can be set according to the risk preference of the maintenance vehicle operator or the uncertainty requirements of the task. If the operator prefers a conservative strategy to avoid potential accidents, the value of δ may be set relatively high. On the contrary, if the task environment is relatively stable and predictable, the value of δ may be relatively low;

[0113] These parameters are determined through the following steps:

[0114] Initial setting: According to the task type of the maintenance vehicle, environmental conditions, and the requirements of the operator, reasonable initial values are set for α t 、β t 、γ t and δ;

[0115] Real-time adjustment: By monitoring the real-time status of the maintenance vehicle and the external environment, such as traffic conditions, weather conditions, etc., the values of these parameters are dynamically adjusted to optimize the performance of the maintenance vehicle;

[0116] Feedback loop: Based on the execution results and performance indicators of the maintenance vehicle, fine-tuning of the parameters is performed to achieve better optimization effects;

[0117] Long-term learning: Over time, through accumulated historical data and machine learning algorithms, the system can learn more effective parameter adjustment strategies to adapt to the changing operating environment and task requirements;

[0118] Preferably, V(P,t) represents the uncertainty value of path P at time t, which reflects the uncertain factors that the path may encounter in the future, such as traffic conditions, weather changes, etc.;

[0119] V(P,t) can be estimated through statistical analysis of historical data and real-time information. For example, based on historical traffic flow data, the possibility of traffic congestion is predicted, or weather forecast data is used to predict the impact of bad weather. In addition, probability models such as Bayesian networks or Markov chains can be used to model and predict future uncertainties;

[0120] Preferably, the data upload process involves steps such as data collection, preprocessing, encrypted transmission, and storage, and the steps are as follows:

[0121] Collect data from sensors;

[0122] Then perform preprocessing steps on the data, such as cleaning, format conversion, and feature extraction;

[0123] Encrypt the data using the SSL / TLS protocol and upload it through the 5G network;

[0124] After the data is decrypted, store it in the cloud data center;

[0125] The data security guarantee measures are as follows:

[0126] Use the TLS / SSL encryption protocol;

[0127] Use algorithms such as CRC, MD5, or SHA to verify the data integrity;

[0128] Use mechanisms such as OAuth and JWT to ensure data access security;

[0129] Deploy a RAID system and back up the data regularly;

[0130] Use firewalls, IPS / IDS systems to prevent unauthorized access;

[0131] Specifically, during the process of data transmission from the 5G network to the cloud data center, security and integrity are mainly ensured in the following ways:

[0132] Encrypted transmission: Use the SSL / TLS protocol or other encryption means to encrypt the data to prevent the data from being eavesdropped or tampered with during transmission;

[0133] Data integrity verification: Verify whether the data remains unchanged during transmission through hash algorithms such as CRC, MD5, or SHA;

[0134] Identity authentication and authorization: Ensure that only authorized devices and users can access the data, and use identity verification mechanisms such as OAuth and JWT;

[0135] Redundancy and backup: Deploy a RAID redundant storage system in the data center and back up the data regularly to prevent data loss;

[0136] It should be noted that the CADPOA algorithm adjusts the optimal path in real time to cope with sudden changes, and immediately uploads the adjusted path to the cloud via the 5G network. The algorithm dynamically optimizes travel time, fuel consumption, and maintenance efficiency, balances certainty and uncertainty, ensures efficient and flexible highway maintenance operations, intelligently adjusts parameters according to task urgency and environmental conditions, combines historical and real-time data to predict future road conditions, improves decision-making accuracy, encrypts data transmission to ensure information security, supports intelligent operation of maintenance vehicles, and improves road maintenance efficiency and safety.

[0137] S4. The highway maintenance vehicle plans its own driving route according to the optimal path instruction issued by the cloud data center and executes the maintenance task.

[0138] Furthermore, after receiving the optimal path instruction issued by the cloud data center, the highway maintenance vehicle first parses the instruction, including the details of the maintenance tasks that the maintenance vehicle needs to execute, the expected driving route, the maintenance demand points along the way, the expected arrival time, as well as safety tips and warning information;

[0139] Based on the parsed instruction, the on-vehicle navigation of the highway maintenance vehicle will automatically plan the driving route;

[0140] The on-vehicle navigation takes into account the current vehicle position, destination, traffic conditions on the expected path, weather forecast, and possible dynamic obstacles;

[0141] And generates a detailed driving plan, including the departure time, expected arrival time, turning points, deceleration areas, acceleration areas, and the staying time at maintenance points;

[0142] The maintenance vehicle executes the maintenance task according to the planned route, and the on-vehicle software monitors the real-time state of the vehicle.

[0143] The real-time state includes speed, fuel consumption, battery power, tire pressure, etc., to ensure that the vehicle runs in the best state. At the same time, the system continuously monitors the vehicle's surrounding environment, such as detecting obstacles through cameras and radars, to ensure driving safety;

[0144] It should be noted that the highway maintenance vehicle autonomously plans the route according to the cloud optimal path instruction, efficiently executes the maintenance, intelligently analyzes the task details, considers real-time traffic and weather, generates an accurate driving plan, ensures arriving at the maintenance point on time, and the vehicle monitors its own state and environment throughout the process to ensure the safety and efficiency of the operation.

[0145] S5. For the highway maintenance vehicles performing tasks on the road, the cloud data center will continuously receive feedback data.

[0146] Furthermore, the cloud data center analyzes in real time the data uploaded by the highway maintenance vehicles performing tasks on the road;

[0147] Dynamically adjust the path planning and evaluate its performance, including travel time, fuel efficiency, safety indicators, path deviation, and task completion rate, to ensure that the maintenance vehicle can quickly respond to new situations, reduce ineffective driving, and improve operation efficiency;

[0148] Travel time: Measure the deviation between the actual travel time and the estimated time;

[0149] Fuel efficiency: Calculate the fuel consumption per unit distance;

[0150] Safety indicators: Evaluate the frequency of safety-related events, such as the number of emergency brakes;

[0151] Path deviation: Compare the difference between the actual driving path and the planned path;

[0152] Task completion rate: Measure the ratio of the number of tasks completed within the specified time to the number of planned tasks;

[0153] When the actual travel time exceeds the expected time by 10%, trigger path replanning. This indicates that the driving time of the maintenance vehicle is longer than planned, which may be due to traffic congestion, poor road conditions, or other unforeseen factors. In this case, triggering path replanning is to find a new and possibly faster route to ensure that the maintenance vehicle can complete the task in a timely manner;

[0154] When the fuel efficiency is less than 12 kilometers per liter of gasoline, adjust the speed or route to save fuel. This means that the fuel consumption of the maintenance vehicle is higher than expected, which may be due to overloading, bad driving habits, or technical problems of the vehicle itself. Adjusting the speed can help improve fuel efficiency and reduce fuel consumption, which is very important for controlling long-term operating costs;

[0155] When the safety indicators deteriorate, take preventive measures, such as reducing speed or choosing a safer alternative route;

[0156] It should be noted that the road maintenance vehicle independently plans the route according to the optimal path instruction from the cloud, efficiently executes the maintenance, the on-vehicle system monitors the real-time status to ensure safe operation, the cloud continuously analyzes and feedbacks data, dynamically optimizes the path, evaluates the performance of travel, fuel, safety, and task completion, automatically adjusts in case of anomalies, greatly improves the operation efficiency and response speed, and realizes intelligent maintenance operations.

[0157] S6. After completing the task, the cloud data center will summarize the operation data of the road maintenance vehicle and conduct in-depth analysis.

[0158] Furthermore, after the road maintenance vehicle completes the task, the cloud data center summarizes the operation data of all road maintenance vehicles, including vehicle status, operation efficiency, environmental factors, road conditions, operation path, safety indicators, and task completion rate;

[0159] Vehicle status: fuel consumption, speed, engine operating status, battery charge (for electric vehicles);

[0160] Operation efficiency: completion time of each maintenance operation, amount of materials used, type of maintenance operation (such as pothole repair, roadblock removal);

[0161] Environmental factors: weather conditions, temperature, humidity, light intensity during operation;

[0162] Road condition information: road type (highway, urban road, etc.), road condition (evenness, wear degree), traffic flow;

[0163] Operation path: comparison of the actual driving path and the planned path, record of the reasons for path changes;

[0164] Safety indicators: number of emergency brakes, number of collision warnings, minimum safety distance from other vehicles;

[0165] Task completion rate: comparison of the actual completion time and the scheduled time for each task, analysis of the reasons for uncompleted tasks;

[0166] Preprocess the aggregated data;

[0167] Preprocessing includes data cleaning, deduplication, outlier detection and handling, missing value imputation, etc., to ensure the accuracy of the analyzed data;

[0168] Construct an analysis framework for in-depth analysis of different performance indicators;

[0169] Use a time series model to identify the efficiency in different time periods, group similar time periods together, and identify the common trend of efficiency;

[0170] Use analysis of variance to test whether there are significant differences in the average efficiency between different operation types, establish a mathematical model between operation type and efficiency, and predict the efficiency of different types of operations;

[0171] Calculate the revenue generated per unit cost, compare the cost-benefit of different operation types and vehicles, explore the relationship between cost and efficiency, and identify opportunities for cost savings;

[0172] Use Geographic Information System (GIS) tools to compare the geographical coordinates of the actual and planned paths and identify deviations;

[0173] Analyze the impact of traffic conditions, weather changes, etc. on path selection by cross-referencing real-time road condition data;

[0174] Use a decision tree model to predict the best path under future road conditions, and learn the strategy of choosing the optimal path under specific conditions by simulating the results of different path selections;

[0175] Using a moving average, identify the trend of safety indicators over time;

[0176] Identify abnormal safety events through statistical tests;

[0177] Use the Kaplan-Meier survival curve to estimate the duration of maintenance effectiveness and identify the time points when the maintenance effectiveness declines;

[0178] Apply the Cox proportional hazards model to explore the factors affecting the persistence of maintenance effectiveness;

[0179] Conduct sentiment analysis on user feedback text to identify positive and negative feedback and understand the impact of maintenance operations on road users;

[0180] Use ARIMA and state space models to predict the probability of vehicle failures;

[0181] Predict the lifespan of vehicle components based on historical maintenance records;

[0182] Use a seasonally adjusted time series model to predict future maintenance requirements;

[0183] Preferably, the time series model is a statistical model used to analyze data points arranged in chronological order, which usually represent the change of a certain variable over time. The time series model can identify trends, seasonal patterns, cyclic fluctuations, and other time-related behaviors in the data. In the context of highway maintenance vehicles, the time series model can be used to identify the operation efficiency during different time periods (such as morning rush hour, evening rush hour, weekends), help plan operation times, and predict future workloads and resource requirements;

[0184] Analysis of variance is a statistical method used to compare whether there are significant differences in the means of two or more groups. In maintenance operations, analysis of variance can be used to test whether there are significant differences in the average operation time or cost between different operation types (such as pavement patching, sign refreshing), and help identify operation types with higher efficiency or better cost-effectiveness;

[0185] GIS tools are software systems used to collect, store, analyze, and display geospatial data. In highway maintenance, GIS can be used to draw and compare the actual driving paths and planned paths of maintenance vehicles, analyze the reasons for path deviations, and can also combine real-time traffic conditions and weather data to evaluate the effectiveness of path selection and support more reasonable path planning;

[0186] Decision tree is a supervised learning algorithm used for classification and regression tasks; in maintenance operations, the decision tree model can predict the optimal path under future road conditions based on historical data. By considering various factors (such as weather, traffic flow, construction areas), the decision tree learns how to select the best path under different conditions;

[0187] Moving average is a statistical method used to analyze time series data. By calculating the average of a series of data points, it can smooth the data and identify trends; in safety index analysis, moving average can help identify the changing trends of safety events (such as the number of emergency brakes) over time, assisting in the monitoring and early warning of safety performance;

[0188] Statistical tests are tools used to determine whether the differences between sample data are statistically significant, such as t-tests, chi-square tests, etc.; in maintenance operations, statistical tests can be used to identify abnormal safety events. For example, if the number of emergency brakes increases abnormally within a certain period, it indicates possible safety hazards;

[0189] Survival curves, especially the Kaplan-Meier curve, are used to describe the survival time distribution of individuals or entities and are usually used in medical research to evaluate treatment effects; in maintenance operations, survival curves can be used to estimate the duration of maintenance effects, identify the long-term effects of maintenance measures, and guide future maintenance strategies;

[0190] Risk models, such as the Cox proportional hazards model, are used to analyze multiple variables that affect survival time; in the field of highway maintenance, risk models can explore various factors that affect the durability of maintenance effects, such as material quality, construction technology, environmental conditions, etc., to help optimize maintenance plans and resource allocation;

[0191] The ARIMA model is a popular time series forecasting method consisting of three parts: autoregression, differencing, and moving average. Its autoregression is that the model uses past observations as predictors for the current value; differencing is to make the time series stationary, and the original data needs to be differenced, that is, subtract the previous observation from each observation; moving average is that the model uses past prediction errors to correct the prediction. The ARIMA model is usually expressed as ARIMA(p, d, q), where p is the number of autoregressive terms, d is the number of differencing times, and q is the number of moving average terms; the ARIMA model is applicable to time series data without obvious seasonal patterns.

[0192] The state space model is a mathematical framework widely used in signal processing, control system theory, and time series analysis. It regards the internal state of the system as an unobservable hidden variable and infers these states through a set of observable outputs. It is divided into two parts: the state equation and the observation equation. The state equation describes how the system state changes over time, while the observation equation describes how the observed data depends on the system state. The state space model is very flexible and can handle complex dynamic systems, including nonlinear and non-stationary processes, and is commonly used in econometrics, engineering control, and other fields.

[0193] Seasonal adjustment time series models are specifically designed to handle data containing periodic patterns. Such models attempt to separate the seasonal component from the data to more accurately analyze trends and cycles. Common seasonal adjustment methods include the multiplicative model and the additive model. The multiplicative model decomposes the time series into trend, seasonal, and random components, assuming that the seasonal effect is proportional to the trend level. The additive model assumes that the seasonal effect is constant and does not change with the trend level. For example, based on the ARIMA model, it can be extended to the seasonal ARIMA model, which not only includes the components of ARIMA but also adds seasonal autoregressive and seasonal moving average parts to better capture seasonal patterns.

[0194] It should be noted that after the task is completed, the cloud data center deeply analyzes the operation data of the maintenance vehicle, covering multiple dimensions such as vehicle status, efficiency, environment, and safety. After preprocessing, the data quality is ensured. Statistical and prediction models are used to evaluate operation efficiency, predict cost-benefit, optimize route selection, monitor safety trends, evaluate maintenance persistence, and understand social impacts through user feedback, predict vehicle health and future needs, comprehensively improving the scientific nature and predictability of road maintenance decisions. The purpose is to achieve optimal allocation of resources, improve maintenance efficiency, ensure road safety, and enhance public satisfaction.

[0195] This embodiment also provides a path planning system for a road maintenance vehicle, including: a data acquisition module, a fusion module, a path calculation module, a prediction and adjustment module, an execution feedback module, and an analysis and optimization module; the data acquisition module is used to collect multi-modal data of the road maintenance vehicle through sensors and a 5G network and upload it to the cloud data center for multi-modal data fusion; the fusion module is used to quickly calculate an optimal path set based on the fused multi-modal data using a quantum optimization algorithm; the path calculation module is used to perform real-time path prediction and adjustment using the CADPOA algorithm based on the calculated optimal path and upload it to the cloud data center using a 5G network; the prediction and adjustment module is used for the road maintenance vehicle to independently plan a driving route and perform maintenance tasks according to the optimal path instruction issued by the cloud data center; the execution feedback module is used for the cloud data center to continuously receive feedback data for the road maintenance vehicle performing tasks on the road; the analysis and optimization module is used to summarize the operation data of the road maintenance vehicle and conduct in-depth analysis after the task is completed.

[0196] This embodiment also provides a computer device applicable to the case of the path planning method for a road maintenance vehicle, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the path planning method for a road maintenance vehicle as proposed in the above embodiment.

[0197] This computer device can be a terminal, and this computer device includes a processor, a memory, a communication interface, a display screen, and an input device connected through a system bus. Among them, the processor of this computer device is used to provide computing and control capabilities. The memory of this computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of this computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be achieved through WIFI, a carrier network, NFC (Near Field Communication), or other technologies. The display screen of this computer device can be a liquid crystal display screen or an electronic ink display screen, and the input device of this computer device can be a touch layer covering the display screen, or a button, a trackball, or a touchpad provided on the housing of the computer device, or an external keyboard, touchpad, or mouse, etc.

[0198] This embodiment also provides a storage medium on which a computer program is stored. When the program is executed by a processor, the path planning method for a highway maintenance vehicle proposed in the above embodiment is implemented; the storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (Static Random Access Memory, referred to as SRAM), electrically erasable programmable read-only memory (Electrically Erasable Programmable Read-Only Memory, referred to as EEPROM), erasable programmable read-only memory (Erasable Programmable Read Only Memory, referred to as EPROM), programmable read-only memory (Programmable Red-Only Memory, referred to as PROM), read-only memory (Read-Only Memory, referred to as ROM), magnetic storage, flash memory, magnetic disk or optical disk.

[0199] In summary, the present invention solves the problems of data fusion, dynamic optimization and intelligent decision-making in the path planning of highway maintenance vehicles by integrating advanced data acquisition, quantum optimization algorithm and real-time path adjustment technology.

[0200] Example 2, referring to Table 1, is the second example of the present invention. To further verify the advancement of the present invention, experimental simulation data of a path planning method for a highway maintenance vehicle are provided.

[0201] The experiment selected two highway maintenance vehicles equipped with similar hardware configurations, marked as the innovative system vehicle and the traditional system vehicle respectively. The experiment was conducted on a typical mixed traffic highway, which includes urban sections, suburban sections and mountainous sections, with a total length of about 200 kilometers.

[0202] First, both vehicles are equipped with the same multimodal sensors, including high-definition cameras, LiDAR, ultrasonic sensors, thermal imaging cameras, GPS / GNSS modules, IMUs, traffic flow monitoring equipment, meteorological sensors, vehicle status monitors, and communication modules. The two vehicles set off at the same time with the goal of completing maintenance work in the designated area, including clearing roadblocks, repairing cracks and potholes, and checking road signs and markings, while ensuring safe driving.

[0203] Next, the innovative system car adopted advanced data collection and quantum optimization algorithms, while the traditional system car used traditional data processing and path planning algorithms. The data collection frequency of the two cars was set to once per second to ensure the real-time and accuracy of the data.

[0204] Finally, the sensor data of the two vehicles are uploaded to the cloud data center in real time for data fusion and path planning. The innovative system vehicle uses the EBFFA algorithm and the QEMOOA algorithm for data fusion and path planning, while the traditional system vehicle uses a rule-based method for data processing and path planning.

[0205] As shown in Table 1 below:

[0206] Table 1 Experimental Record Table

[0207]

[0208] From the above data, it can be seen that the innovative system vehicle adopting the content of the present invention is significantly superior to the traditional system vehicle adopting the traditional path planning method in terms of travel time, fuel consumption, safety score, and task completion rate. Specifically:

[0209] Travel time: The innovative system vehicle only took 120 minutes to complete the maintenance task, while the traditional system vehicle took 150 minutes, which means that the travel time of the innovative system vehicle was shortened by 20%.

[0210] Fuel consumption: The fuel consumption of the innovative system vehicle was 15.5 liters, while the fuel consumption of the traditional system vehicle was 20.2 liters, and the fuel savings reached 23%.

[0211] Safety score: The safety score of the innovative system vehicle was 95, which was much higher than the 88 points of the traditional system vehicle, indicating that the innovative system vehicle can better avoid risks and provide higher safety protection when performing tasks.

[0212] Task completion rate: The task completion rate of the innovative system vehicle reached 98%, while that of the traditional system vehicle was only 90%, which shows that the innovative system vehicle is more efficient when performing maintenance tasks and can more effectively complete the established maintenance work.

[0213] These data fully prove the innovation and advantages of the content of the present invention, especially in improving the operation efficiency of highway maintenance vehicles, reducing operating costs, and enhancing operation safety. By adopting advanced data collection and quantum optimization algorithms, the intelligent highway maintenance vehicle path planning system can achieve better path planning, thus significantly improving the overall efficiency of maintenance operations.

[0214] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered by the scope of the claims of the present invention.

Claims

1. A path planning method for a highway maintenance vehicle, characterized in that: include, Collect multimodal data from road maintenance vehicles through sensors and 5G networks and upload them to the cloud data center for multimodal data fusion; Based on the fused multimodal data, the optimal path set is quickly calculated using quantum optimization algorithms; Based on the calculation of the optimal path, the CADPOA algorithm is used to make real-time path prediction and adjustment and then uploaded to the cloud data center using the 5G network; Highway maintenance vehicles plan their own routes and perform maintenance tasks based on the optimal path instructions issued by the cloud data center; For highway maintenance vehicles performing tasks on the road, the cloud data center will continuously receive feedback data; After completing the task, the cloud data center will summarize the highway maintenance vehicle operation data and conduct in-depth analysis; Based on the fused multimodal data, the QEMOOA algorithm is used to calculate the optimal path at time t, and the expression is: The objective function is expressed as: G＝3； Among them, P * is the optimal path selection value, is to find the path value P, Q that maximizes the expression in the brackets g (P, t) is the quantum expectation value of the g-th objective function under path P and time t, which reflects the optimization ability of quantum computing in probability space. g is the quantum weight factor of the g-th objective function, reflecting the importance of the objective function. g is the index of the objective function, ranging from 1 to G. G is the total number of objective functions. 3 is the three special values ​​in the objective function, corresponding to the total travel time value T(P), the total fuel consumption value E(P) and the total safety value S(P).

2. The path planning method for a highway maintenance vehicle according to claim 1, characterized in that: The multimodal data of the highway maintenance vehicle is collected through sensors and 5G network and uploaded to the cloud data center for multimodal data fusion. The specific steps are as follows: Highway maintenance vehicles collect multimodal data in real time through multimodal sensors; Preprocess the multimodal data; Upload the pre-processed multimodal data to the cloud data center in real time via the 5G network; The EBFFA algorithm is used to fuse the preprocessed data set, and the expression is: Among them, F(D(t)) is the output value of the fusion algorithm, which represents the result of fusion of all data D at time t, and H i (D i (t)) is the i-th sensor data D i The information entropy value of (t) is used to measure the uncertainty and information content of the data, F i (D i (t)) is the i-th sensor data D i The fuzzy membership function value of (t) reflects the relevance and credibility of the data and the fusion target, i is the index of the sensor data, and n is the total number of sensor data.

3. The path planning method for a highway maintenance vehicle according to claim 2, characterized in that: As described above, based on the calculation of the optimal path, the CADPOA algorithm is used to perform real-time path prediction and adjustment and upload it to the cloud data center using the 5G network. The specific steps are as follows: Based on the optimal path, the CADPOA algorithm is used for real-time path adjustment and prediction. The expression is: Among them, P n (t) is the new n-path value after adjustment at time t, It is to find the specific path value that achieves the minimum value among all possible paths, which means to find the path P(t) that minimizes the expression in the brackets among all possible paths P(t), α t is the weight factor for the travel time varying with time t, β t is the weight factor of fuel consumption over time t, γ t is the weight factor of maintenance efficiency that changes with time t, T(P, t) represents the total travel time of path P at time t, E(P, t) represents the total fuel consumption value of path P at time t, C(P, t) represents the maintenance efficiency value of path P at time t, δ is a constant weight factor, and V(P, t) represents the uncertainty value of path P at time t; The adjusted and predicted real-time path is uploaded to the cloud data center via the 5G network.

4. The path planning method for a highway maintenance vehicle according to claim 3, characterized in that: As described above, the highway maintenance vehicle plans its own route and performs maintenance tasks according to the optimal path instructions issued by the cloud data center. The specific steps are as follows: After receiving the optimal route instruction from the cloud data center, the road maintenance vehicle first parses the instruction, including the details of the maintenance task that the maintenance vehicle needs to perform, the expected driving route, the maintenance points along the way, the expected arrival time, and safety tips and warning information; Based on the parsed instructions, the onboard navigation system of the highway maintenance vehicle will automatically plan the driving route; and generates a detailed driving plan including departure time, estimated arrival time, turning points, deceleration zones, acceleration zones, and dwell time at maintenance points; The maintenance vehicle performs maintenance tasks according to the planned route, and the on-board software monitors the real-time status of the vehicle.

5. The path planning method for a highway maintenance vehicle according to claim 4, characterized in that: As mentioned above, for the road maintenance vehicles performing tasks on the road, the cloud data center will continue to receive feedback data. The specific steps are as follows: The cloud data center analyzes data uploaded by highway maintenance vehicles performing tasks on the road in real time; Dynamically adjust path planning and evaluate its performance, including travel time, fuel efficiency, safety indicators, path deviation, and task completion rate; When the actual travel time exceeds the expected time by 10%, path replanning is triggered; Adjust speed when fuel efficiency is less than 12 kilometers per liter of gasoline; When safety indicators deteriorate, preventive measures should be taken.

6. The path planning method for a highway maintenance vehicle according to claim 5, characterized in that: As mentioned above, after completing the task, the cloud data center will summarize the highway maintenance vehicle operation data and conduct in-depth analysis. The specific steps are as follows: After the highway maintenance vehicle completes its mission, the cloud data center aggregates the operation data of all highway maintenance vehicles, including vehicle status, operation efficiency, environmental factors, road condition information, operation path, safety indicators, and mission completion rate; Preprocess the aggregated data; Build an analysis framework to conduct in-depth analysis on different performance indicators; The time series model is used to identify the efficiency in different time periods, and the average efficiency between different job types is tested through variance analysis; Calculate benefits per unit cost and compare the cost-effectiveness of different operation types and vehicles; Use geographic information GIS tools to compare the geographic coordinates of the actual and planned routes, and analyze traffic conditions and weather changes by cross-referencing real-time traffic data; Use decision tree models to predict the best path under future road conditions and use moving averages to identify trends in safety indicators over time; Statistical tests were used to identify unusual safety events, and Kaplan-Meier survival curves were used to estimate the duration of the maintenance effect.

7. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the path planning method for a highway maintenance vehicle according to any one of claims 1 to 6 are implemented.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the path planning method for a highway maintenance vehicle according to any one of claims 1 to 6 are implemented.

Images