Maintenance management system applied to expressway

Abstract

The invention discloses a maintenance management system applied to an expressway, and relates to the technical field of data analysis, and the system comprises the steps: collecting the road state data of the expressway, creating an anomaly detection auto-encoder, marking the abnormal road state data, and building an expressway pavement state trend chart; based on the expressway pavement state trend chart, establishing a pavement degradation prediction model, and predicting a future expressway pavement damage probability distribution diagram; fitting and analyzing the fitness of each executable maintenance scheme based on known maintenance scheme types and the expressway pavement damage probability distribution map, screening out an executable maintenance scheme set of expressway pavement damage at unit time nodes in the future, and taking influence factors during the execution of the maintenance schemes as influence variables; and establishing a dynamic planning state transition equation, and generating an optimal highway maintenance scheme. The method has the advantages that the maintenance cost is reduced, the emergency response speed is increased, and traffic interference is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of data analysis, and in particular to a maintenance management system applied to highways. Background Art

[0002] Highway maintenance management refers to the process of regular inspection, maintenance, repair and updating of highway infrastructure to ensure its safety, smooth operation and normal use.

[0003] Existing highway maintenance management mainly relies on regular maintenance or generates corresponding maintenance plans based on the analysis of road aging status based on the overall external environmental interference. However, there is a mismatch between routine maintenance and the generation of maintenance tasks based on external environmental impact estimation, resulting in poor highway maintenance results, high maintenance costs, and low efficiency. Summary of the Invention

[0004] To solve the above technical problems, a maintenance management system for highways is provided. This technical solution solves the above-mentioned problems that the existing highway maintenance management mainly adopts regular maintenance or generates corresponding maintenance plans based on the analysis of the aging status of the road based on the overall external environmental interference. However, there is a mismatch between the conventional maintenance and the generation scheme of maintenance tasks based on the external environmental impact estimation, resulting in poor highway maintenance results, high maintenance costs and low efficiency.

[0005] In order to achieve the above objects, the technical solution adopted by the present invention is: A maintenance management system applied to a highway, comprising: Abnormal data classification module, pavement degradation prediction module, maintenance plan screening module, maintenance task dynamic planning module; The abnormal data segmentation module collects highway road status data based on distributed edge sensors in each section of the highway, creates an anomaly detection autoencoder, marks abnormal road status data, and constructs a highway pavement status trend map; The pavement degradation prediction module is electrically connected to the abnormal data classification module. The pavement degradation prediction module is used to establish a pavement degradation prediction model based on the highway pavement status trend map and predict the probability distribution map of future highway pavement damage; The maintenance plan screening module is electrically connected to the pavement degradation prediction module. The maintenance plan screening module is used to perform a fitting analysis on the fitness of each executable maintenance plan based on the known maintenance plan type and the highway pavement damage probability distribution map, and screen out a set of executable maintenance plans for highway pavement damage at a future unit time node. The maintenance task dynamic planning module is electrically connected to the maintenance plan screening module. The maintenance task dynamic planning module is used to establish a dynamic planning state transition equation based on a set of executable maintenance plans for highway pavement damage at future unit time nodes, using influencing factors when executing maintenance plans as influencing variables, and generate an optimal highway maintenance plan.

[0006] Preferably, the abnormal data partitioning module specifically includes: A road standardization unit obtains standardized parameters of highway pavement structure and establishes an array of standardized parameters of highway pavement; A data preprocessing unit performs standardization processing on the standardized parameter array of the highway road surface and the highway road status data; a data vector conversion unit, which performs vector conversion between the highway road surface standardized parameter array and the highway road state data according to a linear mapping, and forms a time series vector of the highway road surface state data; The abnormality marking encoder unit, based on the Autoencoder, takes the time series vector of highway road surface state data as input and outputs the abnormal road state data in the highway road surface state data marked per unit time, thereby obtaining the abnormal highway road surface state trend data; A trend visualization unit generates a highway pavement status trend graph based on highway abnormal road pavement status trend data; The Autoencoder-based self-encoder is specifically: , where is the time series vector of highway road surface status data, is the abnormal road state data in the highway road pavement state data per unit time, is the neural network weight matrix, is the regularization coefficient, is the square of the Frobenius norm of the neural network weight matrix.

[0007] Preferably, the pavement degradation prediction module specifically includes: The road section division unit divides and marks the abnormal road pavement condition trend data of each road section in the highway pavement condition trend map, and obtains the abnormal road pavement condition trend characteristic data of each highway section; The road damage prediction unit, based on the ST-GCN model, randomly initializes the weights of the graph convolution layer according to the Gaussian distribution, sets the time convolution layer with hyperparameters, takes the abnormal road pavement status trend feature data of each section of the highway as input, calculates the forward propagation and backpropagation of the graph convolution output and updates the parameters, and obtains the pavement degradation prediction model of each section, and obtains the optimal parameters of the pavement degradation prediction model for each section. ,According to the central aggregation of the global model, the optimal parameters of the pavement degradation prediction model of each road section are used for federal average calculation, the global model parameters are updated, and the training is ended by minimizing the error function of the model to generate the predicted probability of road pavement damage in each section of the highway in the future; The damage visualization unit generates a distribution map of the predicted future highway pavement damage probability based on the predicted future highway pavement damage probability of each section; The pavement degradation prediction model is specifically: , where is the probability of road pavement damage in the i-th section of the highway, is the weight matrix of the fully connected layer, is the feature matrix after the fusion of spatiotemporal features, is the bias term of the fully connected layer; The central aggregation according to the global model is specifically as follows: , where is the global model parameter, is the local parameter of the pavement degradation prediction model for the i-th section of the highway, and N is the total number of sections.

[0008] Preferably, the maintenance plan screening module specifically includes: A maintenance demand unit determines the maintenance demand for the highway pavement damage status at a unit time node based on the highway pavement damage probability distribution map and the highway pavement damage status at a unit time node; The maintenance screening unit, based on the fitness function of the hybrid genetic algorithm, uses the comprehensive maintenance impact factor of the executable maintenance plan corresponding to the maintenance demand of the highway pavement damage state at a unit time node as an interference condition, calculates the fitness index between the maintenance demand of the highway pavement damage state at a unit time node and the known maintenance plan types, and uses time as an association index to establish a set of executable maintenance plans for highway pavement damage at a unit time; the interference conditions include: execution cost, execution efficiency, and execution risk; The fitness function of the hybrid genetic algorithm is specifically: , where is the fitness index between the highway pavement damage state maintenance demand and the Sth known maintenance plan type, is the cost of the Sth known maintenance plan type, is the efficiency of the Sth known maintenance plan type, is the risk of the Sth known maintenance solution type, 、 、 are all weight coefficients.

[0009] Preferably, the maintenance task dynamic planning module specifically includes: The impact factor unit is based on a set of executable maintenance plans for highway pavement damage at future time nodes. The impact factor of each executable maintenance plan on the target road section is marked as an impact variable. The impact variables include maintenance benefits and traffic impact. The system correction unit determines the correction system status of each section of the highway based on the real-time unit time node comprehensive highway data; the correction system status includes: section damage level and traffic flow; The optimal maintenance task unit, based on the influencing factors of each executable maintenance plan on the target road section as the influencing variables and the modified system state of each highway section, establishes a dynamic programming state transition equation to generate the optimal highway maintenance plan for the unit event node highway pavement damage; The dynamic programming state transition equation is specifically: , Where, is the optimal highway maintenance plan S corresponding to the corrected system state of the road pavement damage of the i-th section of the highway under t unit time, The factors affecting the target road section when executing the maintenance plan are used as influencing variables. is the total number of executable maintenance plans, is the corrected system state of the road surface damage on the i-th road segment For the maintenance plan to be executed, the impact factor of the target road section is used as an immediate reward for the influencing variable. is the corrected system state of the road surface damage on the i-th road segment When the maintenance plan is executed, the influencing factors of the target section are used as influencing variables and then transferred to the state The probability of It is the optimal function of the executable maintenance plan S for correcting the system state of the road pavement damage of the i-th section of the highway under t+1 unit time.

[0010] Compared with the prior art, the present invention has the following beneficial effects: This paper proposes a maintenance management solution for highways. Distributed edge sensors collect real-time, full-dimensional highway status data. Autoencoders are used to segment abnormal data and construct road condition trend maps. A spatiotemporal graph convolutional network is used to predict the probability distribution of future road damage. An improved hybrid genetic algorithm is used to select the optimal maintenance plan. Dynamic programming is used to optimize maintenance task scheduling, comprehensively considering cost, efficiency, and traffic impact to generate a globally optimal maintenance plan. The advantages of this solution include reduced maintenance costs, improved emergency response speed, and reduced traffic disruption. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] Figure 1 This is a framework diagram of a maintenance management system used in highways; Figure 2 The internal structure diagram of the abnormal data partitioning module; Figure 3 This is the internal structure diagram of the pavement degradation prediction module; Figure 4 Screening module internal structure diagram for maintenance plan; Figure 5 This is the internal structure diagram of the maintenance task dynamic planning module. DETAILED DESCRIPTION

[0012] The following description is intended to disclose the present invention so that those skilled in the art can implement the present invention. The preferred embodiments described below are merely examples, and those skilled in the art may conceive of other obvious variations.

[0013] Reference Figure 1 As shown, a maintenance management system applied to a highway includes: Abnormal data classification module, pavement degradation prediction module, maintenance plan screening module, maintenance task dynamic planning module; The abnormal data segmentation module collects highway road status data based on distributed edge sensors in each section of the highway, creates an anomaly detection autoencoder, marks abnormal road status data, and constructs a highway pavement status trend map; The pavement degradation prediction module is electrically connected to the abnormal data classification module. The pavement degradation prediction module is used to establish a pavement degradation prediction model based on the highway pavement status trend map and predict the probability distribution map of future highway pavement damage; The maintenance plan screening module is electrically connected to the pavement degradation prediction module. The maintenance plan screening module is used to perform a fitting analysis on the fitness of each executable maintenance plan based on the known maintenance plan type and the highway pavement damage probability distribution map, and screen out a set of executable maintenance plans for highway pavement damage at a future unit time node. The maintenance task dynamic planning module is electrically connected to the maintenance plan screening module. The maintenance task dynamic planning module is used to establish a dynamic planning state transition equation based on a set of executable maintenance plans for highway pavement damage at future unit time nodes, using influencing factors when executing maintenance plans as influencing variables, and generate an optimal highway maintenance plan.

[0014] This solution uses distributed edge sensors to collect full-dimensional highway status data in real time, uses autoencoders to segment abnormal data, and constructs road condition trend maps. It also uses a spatiotemporal graph convolutional network to predict the probability distribution of future road damage. It also uses an improved hybrid genetic algorithm to select the optimal maintenance plan. It also optimizes maintenance task scheduling through dynamic programming, comprehensively considering cost, efficiency, and traffic impact to generate a globally optimal maintenance plan. The advantages of this invention include reduced maintenance costs, improved emergency response speed, and reduced traffic disruption.

[0015] Reference Figure 2 As shown in the figure, the abnormal data partitioning module specifically includes: A road standardization unit obtains standardized parameters of highway pavement structure and establishes an array of standardized parameters of highway pavement; A data preprocessing unit performs standardization processing on the standardized parameter array of the highway road surface and the highway road status data; a data vector conversion unit, which performs vector conversion between the highway road surface standardized parameter array and the highway road state data according to a linear mapping, and forms a time series vector of the highway road surface state data; The abnormality marking encoder unit, based on the Autoencoder, takes the time series vector of highway road surface state data as input and outputs the abnormal road state data in the highway road surface state data marked per unit time, thereby obtaining the abnormal highway road surface state trend data; A trend visualization unit generates a highway pavement status trend graph based on highway abnormal road pavement status trend data; The Autoencoder-based self-encoder is specifically: , where is the time series vector of highway road surface status data, is the abnormal road state data in the highway road pavement state data per unit time, is the neural network weight matrix, is the regularization coefficient, is the square of the Frobenius norm of the neural network weight matrix.

[0016] It is understandable that for the structure of the Autoencoder, the number of hidden layers, the number of neurons, and the activation function, a deep autoencoder with multiple hidden layers can be designed, and the number of neurons in the hidden layers is a tower structure with decreasing or increasing numbers. The activation function can be ReLU or Sigmoid. The division of the training data set (training set, validation set, test set), the selection of loss function (such as mean square error, cross entropy loss, etc.), the selection of optimizer (such as Adam, SGD, etc.), the number of training rounds, batch size, and hyperparameters depend on the training results and are determined by the experience of the implementation technicians. We will not elaborate on this.

[0017] Reference Figure 3 As shown in Figure 1, the pavement degradation prediction module specifically includes: The road section division unit divides and marks the abnormal road pavement condition trend data of each road section in the highway pavement condition trend map, and obtains the abnormal road pavement condition trend characteristic data of each highway section; The road damage prediction unit, based on the ST-GCN model, randomly initializes the weights of the graph convolution layer according to the Gaussian distribution, sets the time convolution layer with hyperparameters, takes the abnormal road pavement status trend feature data of each section of the highway as input, calculates the forward propagation and backpropagation of the graph convolution output and updates the parameters, and obtains the pavement degradation prediction model of each section, and obtains the optimal parameters of the pavement degradation prediction model for each section. ,According to the central aggregation of the global model, the optimal parameters of the pavement degradation prediction model of each road section are used for federal average calculation, the global model parameters are updated, and the training is ended by minimizing the error function of the model to generate the predicted probability of road pavement damage in each section of the highway in the future; The damage visualization unit generates a distribution map of the predicted future highway pavement damage probability based on the predicted future highway pavement damage probability of each section; The pavement degradation prediction model is specifically: , Where, is the probability of road pavement damage in the i-th section of the highway, is the weight matrix of the fully connected layer, is the feature matrix after the fusion of spatiotemporal features, is the bias term of the fully connected layer; The central aggregation according to the global model is specifically as follows: , where is the global model parameter, is the local parameter of the pavement degradation prediction model for the i-th section of the highway, and N is the total number of sections.

[0018] This solution uses distributed edge sensors to collect highway pavement status data in real time, uses autoencoders to classify abnormal data and construct pavement status trend maps. Based on the spatiotemporal graph convolutional network (ST-GCN), the graph convolution layer weights are initialized with a Gaussian distribution, combined with the temporal convolution layer to extract spatiotemporal features, and the model parameters are optimized through forward propagation and backpropagation. Using a federated learning framework, the optimal parameters of the local model for each road section are uploaded to a central server for federated averaging, the global model parameters are updated, the error function is minimized, and ultimately a probability distribution map of future highway pavement damage is generated. The beneficial effect is that it can significantly improve the accuracy of pavement degradation prediction.

[0019] Reference Figure 4 As shown in the figure, the maintenance plan screening module specifically includes: A maintenance demand unit determines the maintenance demand for the highway pavement damage status at a unit time node based on the highway pavement damage probability distribution map and the highway pavement damage status at a unit time node; The maintenance screening unit, based on the fitness function of the hybrid genetic algorithm, uses the comprehensive maintenance impact factor of the executable maintenance plan corresponding to the maintenance demand of the highway pavement damage state at a unit time node as an interference condition, calculates the fitness index between the maintenance demand of the highway pavement damage state at a unit time node and the known maintenance plan types, and uses time as an association index to establish a set of executable maintenance plans for highway pavement damage at a unit time; the interference conditions include: execution cost, execution efficiency, and execution risk; The fitness function of the hybrid genetic algorithm is specifically: , Where, is the fitness index between the highway pavement damage state maintenance demand and the Sth known maintenance plan type, is the cost of the Sth known maintenance plan type, is the efficiency of the Sth known maintenance plan type, is the risk of the Sth known maintenance solution type, 、 、 are all weight coefficients.

[0020] This solution uses a probability distribution map of highway pavement damage to determine the maintenance requirements for pavement damage at specific time points. Integrating an integrated genetic algorithm (IHGA) with comprehensive maintenance influencing factors (such as cost, efficiency, and risk) as interference conditions, it calculates the fitness index of each feasible maintenance plan, selects the optimal set of maintenance plans, and establishes a dynamic maintenance plan using a time index. Implementation significantly improves the adaptability and execution efficiency of maintenance plans, reducing traffic disruptions.

[0021] Reference Figure 5 As shown in the figure, the maintenance task dynamic planning module specifically includes: The impact factor unit is based on a set of executable maintenance plans for highway pavement damage at future time nodes. The impact factor of each executable maintenance plan on the target road section is marked as an impact variable. The impact variables include maintenance benefits and traffic impact. The system correction unit determines the correction system status of each section of the highway based on the real-time unit time node comprehensive highway data; the correction system status includes: section damage level and traffic flow; The optimal maintenance task unit, based on the influencing factors of each executable maintenance plan on the target road section as the influencing variables and the modified system state of each highway section, establishes a dynamic programming state transition equation to generate the optimal highway maintenance plan for the unit event node highway pavement damage; The dynamic programming state transition equation is specifically: , Where, is the optimal highway maintenance plan S corresponding to the corrected system state of the road pavement damage of the i-th section of the highway under t unit time, The factors affecting the target road section when executing the maintenance plan are used as influencing variables. is the total number of executable maintenance plans, is the corrected system state of the road surface damage on the i-th road segment For the maintenance plan to be executed, the impact factor of the target road section is used as an immediate reward for the influencing variable. is the corrected system state of the road surface damage on the i-th road segment When the maintenance plan is executed, the influencing factors of the target section are used as influencing variables and then transferred to the state The probability of It is the optimal function of the executable maintenance plan S for correcting the system state of the road pavement damage of the i-th section of the highway under t+1 unit time.

[0022] A maintenance management system used in highways is as follows: Step 1: Obtain standardized parameters of highway pavement structure and establish a standardized parameter array of highway pavement; Step 2: Standardize the highway pavement standardized parameter array and highway road status data; Step 3: Perform vector conversion between the highway road surface standardized parameter array and the highway road state data according to linear mapping to form a highway road surface state data time series vector; Step 4: Based on the Autoencoder, the time series vector of highway pavement status data is used as input, and the abnormal road state data in the highway pavement status data under the marked unit time is used as output to obtain the abnormal highway pavement status trend data; Step 5: Generate a highway pavement status trend map based on the highway abnormal road pavement status trend data; Step 6: Divide and mark the abnormal road pavement condition trend data of each section in the highway pavement condition trend map to obtain abnormal road pavement condition trend characteristic data of each section of the highway; Step 7: Based on the ST-GCN model, the weights of the graph convolution layer are randomly initialized according to the Gaussian distribution, and the temporal convolution layer is set with hyperparameters. The abnormal road pavement condition trend characteristic data of each section of the highway is used as input, and the forward propagation and backpropagation of the graph convolution output and the parameter update are calculated to obtain the road degradation prediction model of each section and the optimal parameters of each road section road degradation prediction model are obtained. ,According to the central aggregation of the global model, the optimal parameters of the pavement degradation prediction model of each road section are used for federal average calculation, the global model parameters are updated, and the training is ended by minimizing the error function of the model to generate the predicted probability of road pavement damage in each section of the highway in the future; Step 8: Generate a predicted future highway pavement damage probability distribution map based on the predicted future highway pavement damage probability of each section; Step 9: Based on the highway pavement damage probability distribution map, determine the highway pavement damage status maintenance requirements per unit time node based on the highway pavement damage status per unit time node; Step 10: Based on the fitness function of the hybrid genetic algorithm, the comprehensive maintenance impact factor of the executable maintenance plan corresponding to the maintenance demand of the highway pavement damage state at a unit time node is used as the interference condition. The fitness index between the maintenance demand of the highway pavement damage state at a unit time node and the known maintenance plan type is calculated. Time is used as the association index to establish a set of executable maintenance plans for highway pavement damage at a unit time. Step 11: Based on the set of executable maintenance plans for highway pavement damage at future time nodes, mark the influencing factors of each executable maintenance plan on the target road section as influencing variables; Step 12: Determine the correction system status of each section of the highway based on the real-time unit time node highway comprehensive data; Step 13: Based on the influencing factors of each executable maintenance plan on the target road section as influencing variables and the modified system states of each highway section, a dynamic programming state transition equation is established to generate the optimal highway maintenance plan for the unit event node highway pavement damage.

[0023] The above shows and describes the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions merely illustrate the principles of the present invention. Various changes and modifications may be made to the present invention without departing from the spirit and scope of the present invention. Such changes and modifications are intended to fall within the scope of the present invention. The scope of protection claimed by the present invention is defined by the appended claims and their equivalents.

Claims

1. A maintenance management system applied to highways, characterized in that: include: Abnormal data classification module, pavement degradation prediction module, maintenance plan screening module, maintenance task dynamic planning module; The abnormal data segmentation module collects highway road status data based on distributed edge sensors in each section of the highway, creates an anomaly detection autoencoder, marks abnormal road status data, and constructs a highway pavement status trend map; The pavement degradation prediction module is electrically connected to the abnormal data classification module. The pavement degradation prediction module is used to establish a pavement degradation prediction model based on the highway pavement status trend map and predict the probability distribution map of future highway pavement damage; The maintenance plan screening module is electrically connected to the pavement degradation prediction module. The maintenance plan screening module is used to perform a fitting analysis on the fitness of each executable maintenance plan based on the known maintenance plan type and the highway pavement damage probability distribution map, and screen out a set of executable maintenance plans for highway pavement damage at a future unit time node. The maintenance task dynamic planning module is electrically connected to the maintenance plan screening module. The maintenance task dynamic planning module is used to establish a dynamic planning state transition equation based on a set of executable maintenance plans for highway pavement damage at future unit time nodes, using influencing factors when executing maintenance plans as influencing variables, and generate an optimal highway maintenance plan.

2. A maintenance management system for highways according to claim 1, characterized in that: The abnormal data partitioning module specifically includes: A road standardization unit obtains standardized parameters of highway pavement structure and establishes an array of standardized parameters of highway pavement; A data preprocessing unit performs standardization processing on the standardized parameter array of the highway road surface and the highway road status data; a data vector conversion unit, which performs vector conversion between the highway road surface standardized parameter array and the highway road state data according to a linear mapping, and forms a time series vector of the highway road surface state data; The abnormality marking encoder unit, based on the Autoencoder, takes the time series vector of highway road surface state data as input and outputs the abnormal road state data in the highway road surface state data marked per unit time, thereby obtaining the abnormal highway road surface state trend data; The trend visualization unit generates a highway pavement status trend graph based on the highway abnormal road pavement status trend data.

3. A maintenance management system for highways according to claim 2, characterized in that: The Autoencoder-based self-encoder is specifically: , where is the time series vector of highway road surface status data, is the abnormal road state data in the highway road pavement state data per unit time, is the neural network weight matrix, is the regularization coefficient, is the square of the Frobenius norm of the neural network weight matrix.

4. A maintenance management system for highways according to claim 3, characterized in that: The pavement degradation prediction module specifically includes: The road section division unit divides and marks the abnormal road pavement condition trend data of each road section in the highway pavement condition trend map, and obtains the abnormal road pavement condition trend characteristic data of each highway section; The road damage prediction unit, based on the ST-GCN model, randomly initializes the weights of the graph convolution layer according to the Gaussian distribution, sets the time convolution layer with hyperparameters, takes the abnormal road pavement status trend feature data of each section of the highway as input, calculates the forward propagation and backpropagation of the graph convolution output and updates the parameters, and obtains the pavement degradation prediction model of each section, and obtains the optimal parameters of the pavement degradation prediction model for each section. ,According to the central aggregation of the global model, the optimal parameters of the pavement degradation prediction model of each road section are used for federal average calculation, the global model parameters are updated, and the training is ended by minimizing the error function of the model to generate the predicted probability of road pavement damage in each section of the highway in the future; The damage visualization unit generates a predicted future highway pavement damage probability distribution map based on the predicted future highway pavement damage probability of each section.

5. A maintenance management system for highways according to claim 4, characterized in that: The pavement degradation prediction model is specifically as follows: , where is the probability of road pavement damage in the i-th section of the highway, is the weight matrix of the fully connected layer, is the feature matrix after the fusion of spatiotemporal features, is the bias term of the fully connected layer; The central aggregation according to the global model is specifically as follows: , where is the global model parameter, is the local parameter of the pavement degradation prediction model for the i-th section of the highway, and N is the total number of sections.

6. A maintenance management system for highways according to claim 5, characterized in that: The maintenance plan screening module specifically includes: A maintenance demand unit determines the maintenance demand for the highway pavement damage status at a unit time node based on the highway pavement damage probability distribution map and the highway pavement damage status at a unit time node; The maintenance screening unit, based on the fitness function of the hybrid genetic algorithm, uses the comprehensive maintenance impact factor of the executable maintenance plan corresponding to the maintenance demand of the highway pavement damage status at a unit time node as an interference condition, calculates the fitness index between the maintenance demand of the highway pavement damage status at a unit time node and the known maintenance plan type, and uses time as an association index to establish a set of executable maintenance plans for highway pavement damage under unit time; the interference conditions include: execution cost, execution efficiency, and execution risk.

7. A maintenance management system for highways according to claim 6, characterized in that: The fitness function of the hybrid genetic algorithm is specifically: , where is the fitness index between the highway pavement damage state maintenance demand and the Sth known maintenance plan type, is the cost of the Sth known maintenance plan type, is the efficiency of the Sth known maintenance plan type, is the risk of the Sth known maintenance solution type, 、 、 are all weight coefficients.

8. A maintenance management system for highways according to claim 7, characterized in that: The maintenance task dynamic planning module specifically includes: The impact factor unit is based on a set of executable maintenance plans for highway pavement damage at future time nodes. The impact factor of each executable maintenance plan on the target road section is marked as an impact variable. The impact variables include maintenance benefits and traffic impact. The system correction unit determines the correction system status of each section of the highway based on the real-time unit time node comprehensive highway data; the correction system status includes: section damage level and traffic flow; The optimal maintenance task unit establishes a dynamic programming state transfer equation based on the influencing factors of the target section for each executable maintenance plan as the influencing variables and the corrected system states of each section of the highway, and generates the optimal highway maintenance plan for the damaged highway pavement per unit event node.

9. A maintenance management system for highways according to claim 8, characterized in that: The dynamic programming state transfer equation is specifically: , Where, is the optimal highway maintenance plan S corresponding to the corrected system state of the road pavement damage of the i-th section of the highway under t unit time, The factors affecting the target road section when executing the maintenance plan are used as influencing variables. is the total number of executable maintenance plans, is the corrected system state of the road surface damage on the i-th road segment For the maintenance plan to be executed, the impact factor of the target road section is used as an immediate reward for the influencing variable. is the corrected system state of the road surface damage on the i-th road segment When the maintenance plan is executed, the influencing factors of the target section are used as influencing variables and then transferred to the state The probability of It is the optimal function of the executable maintenance plan S for correcting the system state of the road pavement damage of the i-th section of the highway under t+1 unit time.

Images