Physical data dual-drive road maintenance decision-making system

Abstract

The invention relates to a physical data dual-drive road maintenance decision-making system, which performs stake mark-road section-road network three-stage maintenance, and can provide a stake mark-level maintenance scheme, a road section-level maintenance scheme and a road network-level maintenance scheme with the precision of 100 meters. According to the maintenance schemes of all levels, the cost, carbon emission, construction time and pavement use performance of road maintenance construction can be considered at the same time, and Pareto optimization among all targets is achieved, so that the final maintenance schemes of all levels can guarantee the optimal pavement use performance while saving the construction cost, reducing the construction carbon emission and reducing the construction time, and the construction efficiency is improved. And the optimal maintenance effect is achieved. According to the three-stage maintenance scheme, the most time-consuming stage does not need to use PINN prediction, and the combination of different maintenance schemes of the Pareto optimal maintenance scheme set provided by the next stage is directly used, so that a large amount of time is saved.

Description

Technical Field

[0001] The present invention relates to the technical field of road maintenance, and in particular to a road maintenance decision system driven by both physical data and physical data. Background Art

[0002] Building an increasingly perfect scientific decision-making management and technical system for highway maintenance and studying the implementation path of scientific maintenance decision-making driven by both physical data are not only important requirements of relevant superior departments for maintenance management, but also rigid demands of maintenance management agencies at all levels in the industry. Therefore, it is very important to establish a scientific, reasonable, sustainable and implementable maintenance decision-making model.

[0003] The shortcomings of existing technologies are that traditional road maintenance work is too dependent on the subjective experience of maintenance engineers, while current advanced maintenance technologies lack consideration of real physical constraints and cannot comprehensively consider conditions such as cost, carbon emissions, long-term road performance, and construction time. Data-based maintenance decisions, especially those based on deep learning, cannot guarantee the correct consideration of the impact of objective factors in the maintenance process, such as traffic factors, geological factors, weather and environmental factors, road structure, maintenance methods, etc., which in turn affects the reliability of maintenance decision results. Summary of the invention

[0004] In view of the shortcomings of existing maintenance technologies, the technical problem to be solved by the present invention is to provide a road maintenance decision system driven by both physical data and physical data. The system can fully and effectively consider the impact of traffic factors, geological factors, weather and environmental factors, road structure, maintenance methods, etc. on the performance of road surfaces, forming a three-stage maintenance decision system of pile number-road section-road network, and can provide pile number-level maintenance plans, road section-level maintenance plans, and road network-level maintenance plans with an accuracy of 100 meters.

[0005] In order to solve the above technical problems, the technical solution of the present invention is: The present invention provides a road maintenance decision system driven by both physical data and physical data, the system comprising: The pavement performance prediction module is used to predict the pavement performance in the next year based on the road age, traffic factors, geological factors, weather and environmental factors, road structure, maintenance methods and the pavement performance values ​​in the past year; The pile number level fitness calculation module is used to calculate the cost, carbon emission, construction time and comprehensive pavement performance of each pile number according to the maintenance plan output by the pile number level optimization module and the pavement performance predicted by the pavement performance prediction module as the pile number level fitness; The pile number optimization module uses a multi-objective exponential distribution optimization algorithm to perform multi-objective optimization on the pile number fitness calculated by the pile number fitness calculation module, outputs a maintenance plan, and obtains the pile number Pareto optimal maintenance plan set for each pile number after reaching the number of iterations. The pile number Pareto optimal maintenance plan set includes G optimal maintenance plans; A section-level fitness calculation module, used to calculate the cost, carbon emission, construction time and comprehensive pavement performance of each section according to the Pareto optimal maintenance plan set of the pile number and the comprehensive pavement performance of each pile number obtained by the pile number-level fitness calculation module, as the section-level fitness; The section-level optimization module inputs the Pareto optimal maintenance plan set of each pile number, and uses the multi-objective exponential distribution optimization algorithm to perform multi-objective optimization on the section-level fitness calculated by the section-level fitness calculation module, outputs the maintenance plan, and obtains the Pareto optimal maintenance plan set of each section after reaching the number of iterations; A road network-level fitness calculation module, used to calculate the cost, carbon emission, construction time and comprehensive road performance of the road network according to the Pareto optimal maintenance plan set of the road section and the comprehensive road performance of each road section obtained by the road section-level fitness calculation module, as the road network-level fitness; The road network level optimization module takes the Pareto optimal maintenance plan set of the road section as input, uses the multi-objective exponential distribution optimization algorithm to perform multi-objective optimization on the road network level fitness calculated by the road network level fitness calculation module, outputs the maintenance plan, and obtains the road network Pareto optimal maintenance plan set after reaching the number of iterations.

[0006] Furthermore, the pavement performance prediction module uses a physical embedded neural network PINN for prediction, including an input layer, a hidden layer, an output layer and a loss term; The loss item It consists of four parts, and the calculation formula is: (19) in, Represents all parameters of the physical embedded neural network, is the data loss term coefficient, is the data loss term, is the physical loss term coefficient, is the physical loss term, is the boundary loss term coefficient, is the boundary loss term, is the regular loss term coefficient, is the regular loss term.

[0007] Furthermore, in the pile number level fitness calculation module, the calculation formulas for the cost Cost, carbon emission CE, construction time CT and comprehensive pavement performance CRPS of each pile number are as follows: (4) (5) (6) (7) in, It represents the cost of each maintenance plan corresponding to each pile number to adopt the g-th maintenance method in the tth year, t0 represents the starting year, t e Indicates the year of the end; It represents the carbon emission of each maintenance scheme corresponding to each pile number when the g-th maintenance method is adopted in the t-th year; It indicates the time required for each maintenance scheme corresponding to each pile number to be constructed by the g-th maintenance method in the t-th year; the value of g is an integer from 1 to 9, representing micro-surfacing, pavement lining, thermal regeneration, single-layer excavation and patching, single-layer overlay, double-layer excavation and patching, double-layer overlay, triple-layer excavation and patching, and roadbed grouting respectively; Represents the physical embedded neural network in the input vector The predicted value below.

[0008] Furthermore, in the road section-level fitness calculation module, the calculation formulas for the cost CL, carbon emission CEL, comprehensive pavement performance CRPSL and construction time CTL of each road section are as follows: (26) (27) (28) (29) Where n represents the number of pile numbers; h represents the hth 100-meter pile number in the road section; represents an optimal maintenance plan in the Pareto optimal maintenance plan set for the pile number h; It represents an optimal maintenance plan in the Pareto optimal maintenance plan set for the pile number h+1; , The optimal maintenance plan for the section from pile number h to pile number h+1 is arrive Additional costs and construction time; , , , They represent the optimal maintenance plan M of the pile number h in the Pareto optimal maintenance plan set. h Cost, carbon emissions, pavement performance, and construction time.

[0009] Furthermore, the traffic factors include the average annual passenger car traffic volume, passenger-to-cargo ratio, average annual truck traffic volume and equivalent load times; the geological factors include base properties, geological structure and soil characteristics; the weather environment factors include rainfall, annual average daily maximum temperature and annual average daily minimum temperature; the road structure includes surface layer materials, surface layer thickness, base layer materials and base layer thickness; the maintenance methods include micro-surfacing, pavement patching, thermal regeneration, single-layer excavation and patching, single-layer overlay, double-layer excavation and patching, double-layer overlay, triple-layer excavation and patching and roadbed grouting.

[0010] Furthermore, the multi-objective exponential distribution optimization algorithm in the pile level optimization module includes the following process: initializing a population P, the number of individuals in the population is K, and each individual is an N×M maintenance plan matrix, where N is the length of the maintenance year, M is the number of maintenance methods, and each individual in population P Represents the N-year maintenance plan of the pile number, where each individual maintenance plan satisfies that only one maintenance method is adopted each year, and each row of the maintenance plan matrix satisfies the constraint shown in formula (9): (9) Initialize the individual according to formula (10): (10) Among them, rand(1,M) represents randomly selecting an index column from {1,2,…,M} according to exponential distribution; The exploration and exploitation principle is used to calculate the guided solution according to formula (11). : (11) Among them, time represents the number of iterations, , , Represents the three best solutions at the current number of iterations; The output results of the pile level fitness calculation module are used for information feedback. The update logic of each individual maintenance plan is as follows: the weighted coefficients of the guided solution and the memoryless solution are obtained based on the pile level fitness to calculate the value of the new individual matrix. , and The maximum value of each row is assigned to 1, and the other values ​​are assigned to 0. Then the next generation of individuals is updated according to formula (12): (12) in, It represents selecting the index column j with the largest value for the i-th row of the current matrix, and the corresponding position of the index column with the largest value is assigned 1; is the updated individual of the i-th row, j-th column and time+1-th iteration; is defined as: (13) in, is the memoryless solution of the i-th row and j-th column in the time-th generation individual; a and b are weighted coefficients, which are defined as: (14) in, is the fitness of the guiding solution, represents the fitness of the kth individual, and both are obtained by the pile-level fitness calculation module.

[0011] Furthermore, the pile number-level fitness calculation module, the section-level fitness calculation module, and the road network-level fitness calculation module all use the maintenance specification recommendations as restriction conditions. When the pavement performance value is below 90, micro-surfacing will not be adopted, and when the pavement performance value is below 80, covering measures will not be adopted.

[0012] Compared with the prior art, the present invention has the following beneficial effects: 1. The system of the present invention is a three-stage maintenance decision system of pile number-road section-road network, which can provide pile number-level maintenance plan, road section-level maintenance plan and road network-level maintenance plan with an accuracy of 100 meters. The maintenance plans at all levels can simultaneously consider the cost, carbon emissions, construction time and pavement performance of road maintenance construction, and achieve the Pareto optimality among various goals, so that the final maintenance plans at all levels can save construction costs, reduce construction carbon emissions, and reduce construction time while ensuring the optimal pavement performance and achieving the best maintenance effect.

[0013] 2. The present invention uses the multi-objective optimization algorithm for multiple times to perform multi-level optimization, and the optimization of the pile number-road section-road network in stages can significantly improve the optimization efficiency. The pile number level optimization module sets the search space of the multi-objective exponential distribution algorithm to an N×M matrix, which is used to represent the maintenance plan for many years. In the population initialization process, the restriction of only performing maintenance once a year is added. In the process of updating the next generation population, the algorithm is revised to: calculate the weight through the fitness value and update the maintenance plan matrix, select the largest value in each row and assign it to 1, and assign the others to 0. The road section level optimization module and the road network level optimization module set the search space of the multi-objective exponential distribution algorithm to a vector composed of the pile number Pareto optimal maintenance plan set of each pile number or a vector composed of the road section Pareto optimal maintenance plan set of each road section, and keep the individual vectors in the population as integer vectors during the population initialization process and the process of updating the next generation population. The limitations in actual maintenance construction (each 100-meter pile number is only maintained once a year) are fully considered, and the exponential distribution probability enables the search space of the algorithm to be effectively explored, improves the optimization efficiency, and avoids falling into the local optimum.

[0014] 3. The present invention adopts PINN to integrate physical constraints. In the pavement performance prediction module, the present invention converts the influence mechanism of factors such as road age, traffic factors (average annual passenger car traffic volume, passenger-cargo ratio, average annual truck traffic volume, equivalent load times), geological factors (base properties, geological structure, soil characteristics), weather and environmental factors (rainfall, average annual daily maximum temperature, average annual daily minimum temperature), road structure (surface material, surface thickness, base material, base thickness), maintenance methods (micro-surfacing, pavement patching, thermal regeneration, single-layer excavation, single-layer overlay, double-layer excavation, double-layer overlay, three-layer excavation, roadbed grouting) on ​​pavement performance and maintenance effect into partial differential equations, and then uses the partial differential equations as loss term constraints of the physical embedded neural network. The above-mentioned influencing factors combined with the pavement performance values ​​of the past year are used as input, and the pavement performance values ​​of the next year are used as output to train the physical embedded neural network (PINN). In the fitness calculation, the additional construction time and construction cost required for switching maintenance methods between adjacent 100-meter pile numbers during the maintenance process are taken into account, as well as the maintenance specification recommendations. The two are used as constraints. At the same time, the upper limit of cost, carbon emissions, and construction time during the construction process and the lower limit of pavement performance are considered. The physical constraints are effectively integrated into a whole, which improves the accuracy and stability of prediction and decision-making.

[0015] 4. The three-stage maintenance plan of the present invention will make the most time-consuming stage directly use the combination of different maintenance plans of the Pareto optimal maintenance plan set provided by the next level without using PINN prediction (each maintenance plan combination corresponds to a different fitness, so the next level maintenance plan strategy is directly arranged and combined without complex PINN prediction), saving a lot of time. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 The figure is a flow chart of a multi-objective exponential distribution optimization algorithm MOEDO according to an embodiment of the present invention.

[0017] Figure 2 A schematic diagram of the structure of a physically embedded neural network according to an embodiment of the present invention.

[0018] Figure 3 The present invention is a schematic structural diagram of a three-stage maintenance decision system of pile number-road section-road network according to an embodiment of the present invention. DETAILED DESCRIPTION

[0019] The following are specific embodiments and drawings of the present invention, which are only used to further explain the present invention in detail, but do not limit the protection scope of the present application. It should be noted that the embodiments and features in the embodiments of the present invention can be combined with each other without conflict.

[0020] A road section is a traffic route between two adjacent nodes on a traffic network, and a road network refers to a road system composed of various roads that are interconnected and interwoven into a network distribution within a certain area. The present invention constructs a three-stage maintenance decision system of pile number-road section-road network, which can calculate the road section and even road network-level maintenance plan with an accuracy of 100-meter pile number, avoiding the problem of falling into a local optimal solution due to the existence of a large number of calculations.

[0021] The present invention fully and effectively considers the influence of factors such as road age, traffic factors (average annual traffic volume of passenger cars, passenger-cargo ratio, average annual traffic volume of trucks, equivalent load times), geological factors (base properties, geological structure, soil characteristics), weather and environmental factors (rainfall, average annual daily maximum temperature, average annual daily minimum temperature), road structure (surface material, surface thickness, base material, base thickness) and maintenance methods (micro-surfacing, pavement patching, thermal regeneration, single-layer excavation, single-layer overlay, double-layer excavation, double-layer overlay, three-layer excavation, roadbed grouting) on ​​the pavement performance. The present invention makes full use of the multi-objective optimization algorithm MOEDO in combination with the maintenance requirements of the pile level, section level and road network level, adopts non-dominated sorting and congestion distance mechanism, can effectively balance the conflicts between multiple objectives, and has the advantages of fast convergence speed and not easy to fall into the local optimal solution, and takes the restrictions recommended by the maintenance specifications in the actual maintenance project as the optimization conditions in the optimization process, can provide a more reliable and more feasible maintenance method, and provide a more guiding maintenance plan for the maintenance project. In the fitness calculation, the cost, carbon emissions, construction time and comprehensive pavement performance of road maintenance construction at different levels are considered simultaneously. The optimal comprehensive pavement performance is obtained while saving construction costs, reducing construction carbon emissions and reducing construction time, achieving the Pareto optimality among various goals and achieving a better maintenance effect. The maintenance effect refers to the change in pavement performance after maintenance. If the pavement performance continues to remain at a high level after maintenance, the maintenance effect is good. If the pavement performance decreases rapidly after maintenance, the maintenance effect is poor.

[0022] In the embodiment of the present invention, a physical embedded neural network PINN is constructed for prediction, which fully considers the influence of the above-mentioned influencing factors on pavement performance and maintenance effect, converts them into partial differential equations, and uses the partial differential equations as constraints of the physical embedded neural network to train a pavement performance prediction model that conforms to the actual physical laws.

[0023] The present invention comprises three modules: an optimization module, a pavement performance prediction module and a fitness calculation module. The optimization module is divided into a pile number level optimization module, a section level optimization module and a road network level optimization module, and each optimization module uses a multi-objective optimization algorithm to output a Pareto optimal maintenance plan set of each level. The fitness calculation module is divided into a pile number level fitness calculation module, a section level fitness calculation module and a road network level fitness calculation module, which respectively calculate the cost, carbon emission, construction time and comprehensive pavement performance according to maintenance plans of different levels, and use the four as fitness. The pavement performance prediction module is used to predict the pavement performance in the next year based on the road age, traffic factors, geological factors, weather and environmental factors, road structure, maintenance methods and pavement performance values ​​of the past year. Among them, the optimization module adopts a multi-objective exponential distribution optimization algorithm, including the initialization, exploration, utilization and update of the population, etc., which all use the exponential distribution. Figure 1 This is a flowchart of the multi-objective exponential distribution optimization algorithm. The specific process is: first, initialize the population according to the exponential distribution, initialize the parameters to create an initial population containing K random solutions, generate the child population from the parent population, set k=1, use the information feedback mechanism IFM to update the individuals, add the updated individuals to the population until K individuals are obtained, merge the parent and child populations, perform non-dominated sorting and divide the non-dominated levels, select individuals with high dominance levels, complete the population number according to the crowding sorting, check whether the number of iterations has been reached, if so, end the iteration and output the result to obtain the Pareto optimal solution, if not, continue the iteration.

[0024] The exponential distribution function f(x) in the present invention follows the following formula: (1) in, is the exponential distribution parameter; x represents the input; but The cumulative distribution function F(x) is: (2) In order to prevent the algorithm from falling into the local optimal solution, the memoryless probability distribution P is introduced. The specific formula is as follows: (3) Among them, s represents the time of the event that has occurred, and t represents the length of time that has passed since time s. This formula shows that the probability of the future is only related to the current time period t, and has nothing to do with the time s that has passed.

[0025] Embodiment 1: In this embodiment, the input of the pile number level optimization module is the pile number level fitness corresponding to the pile number, and the output is the maintenance plan of each pile number. After the iteration is completed, the pile number Pareto optimal maintenance plan set is output, and the pile number Pareto optimal maintenance plan set contains multiple optimized maintenance plans, that is, multiple optimal maintenance plans. The pile number level fitness includes the cost, carbon emissions, construction time and comprehensive pavement performance of each pile number, which are obtained through the pile number level fitness calculation module.

[0026] In the pile number level fitness calculation module, the calculation formulas for the cost Cost, carbon emission CE, construction time CT and comprehensive pavement performance CRPS of each pile number are as follows: (4) (5) (6) (7) in, It represents the cost of each maintenance plan corresponding to each pile number to adopt the g-th maintenance method in the tth year, t0 represents the starting year, t e Indicates the year of the end; It represents the carbon emission of each maintenance scheme corresponding to each pile number when the g-th maintenance method is adopted in the t-th year; It indicates the time required for each maintenance scheme corresponding to each pile number to adopt the g-th maintenance method in the t-th year; the value of g is an integer from 1 to 9, representing nine maintenance methods: micro-surfacing, pavement lining, thermal regeneration, single-layer excavation and patching, single-layer overlay, double-layer excavation and patching, double-layer overlay, triple-layer excavation and patching, and roadbed grouting. Represents the physical embedded neural network in the input vector The predicted value under the condition is the output of the road performance prediction module; the input vector It refers to the road age, traffic factors, geological factors, weather and environmental factors, road structure, maintenance methods and road performance values ​​in the past year.

[0027] The pile number level optimization module uses a multi-objective exponential distribution optimization algorithm to perform multi-objective optimization on the pile number level fitness calculated by the pile number level fitness calculation module, outputs a maintenance plan, and obtains the pile number Pareto optimal maintenance plan set for each pile number after reaching the number of iterations. The pile number Pareto optimal maintenance plan set includes G optimal maintenance plans; the number of optimal maintenance plans G contained in the pile number Pareto optimal maintenance plan set is determined by the non-dominated sorting and congestion selection results, and each optimal maintenance plan is calculated separately when performing fitness calculations.

[0028] The multi-objective exponential distribution optimization algorithm in the pile-level optimization module first initializes the population according to the exponential distribution, generates a new generation of population according to formula (8), merges the parent generation and the child generation, performs non-dominated sorting and divides the non-dominated level, selects individuals with high dominance level, completes the population number according to the crowding sorting, and checks whether the number of iterations has been reached. If it has been reached, the iteration is terminated and the result is output. If not, the iteration is continued. It includes the following processes: Initialize the population P, the number of individuals in the population is K, and each individual is an N×M matrix, where N represents the length of the maintenance year and M represents the number of maintenance methods. The initial population is expressed by formula (8): (8) Each individual in population P represents the N-year maintenance plan of the pile number, where each individual maintenance plan needs to satisfy the requirement of only one maintenance method per year. Each individual represents a solution, that is, each row of the matrix satisfies the constraint shown in formula (9): (9) Among them, M is the type of maintenance method, It represents the value of the kth individual in the i-th row and j-th column. When each individual in the population is initialized, it still needs to meet the requirement of only using one maintenance method each year, and is initialized according to the following formula: (10) Here, rand(1,M) represents randomly selecting an index column from {1,2,…,M} according to exponential distribution.

[0029] The exploration and exploitation principle is used to calculate the guided solution (current optimal solution) according to formula (11): : (11) Among them, time represents the number of iterations, , , Represents the three best solutions under the current number of iterations; The output results of the pile-level fitness calculation module are used for information feedback, which is called the information feedback mechanism (IFM). The update logic of each individual maintenance plan is: the pile-level fitness of the pile-level fitness calculation module is used as the weighted coefficient of the guide solution and the memoryless solution to calculate the value of the new individual matrix , and assign the maximum value of each row to 1, and the other values ​​to 0, to ensure that the constraint of only one maintenance method is adopted each year. The next generation of individuals is updated according to the following formula: (12) Among them, It represents selecting the index column j with the largest value for the i-th row of the current matrix, and the corresponding position of the index column with the largest value is assigned 1; is the updated individual of the i-th row, j-th column and time+1-th iteration; is defined as follows: (13) in is the memoryless solution of the i-th row and j-th column in the time-th generation individual; a and b are weighted coefficients, which are defined as: (14) in is the fitness of the guiding solution, Represents the fitness of the kth individual. Both fitnesses are obtained by the pile number-level fitness calculation module. There are four fitness values, namely cost, carbon emission, construction time and comprehensive pavement performance. Their respective weighted coefficients are obtained for multi-objective optimization.

[0030] Embodiment 2: The road section optimization module and the road network optimization module of this embodiment use similar multi-objective exponential distribution optimization algorithms for optimization, with the input being the Pareto optimal maintenance solution set output by the next-level optimization module, and the output being the Pareto optimal maintenance solution set for each road section or the Pareto optimal maintenance solution set for the road network. The specific process is: Assume the size of the initial population is K, Represents an n-dimensional integer vector, where n represents the number of hundred-meter piles that make up the road section or the number of road sections that make up the road network: (15) Each individual in the population is initialized according to the following formula, which means that in the range Generate a random integer randomly according to exponential distribution: (16) in They represent the upper and lower bounds of the h-th variable respectively, h represents the h-th hundred-meter stake number that constitutes the road section or the h-th road section that constitutes the road network; randint represents the random exponential distribution function.

[0031] The update logic of the individual maintenance plan is: if the current individual is a memoryless solution individual, it is updated according to the guided solution and the memoryless solution; if the current solution is not a memoryless solution and the random number d is less than 0.5, it is updated according to the memoryless solution and the current solution; otherwise, it is updated according to the average of all solutions and the current solution. The formula is: (17) Among them, the values ​​of a and b are the same as those in formula (14), and they are obtained by using the corresponding fitness calculated by the road segment level fitness calculation module or the road network level fitness calculation module; round() represents rounding up to an integer value; , is a random value of (0,1), time is the current number of iterations, Max_time is the maximum number of iterations, is the average value of all solutions in the current generation, is the guiding solution of the current generation, is the memoryless solution of the k-th individual of the time-th generation, is the time-th generation k-th solution. is the variance, is the exponential distribution parameter, is a random parameter greater than 1, and d is a random parameter between 0 and 1; is the kth updated individual of the time+1th generation. , is defined as: (18) and Plots the average of all solutions and two individuals randomly selected from the initial population and The distance between.

[0032] The multi-objective exponential distribution optimization algorithm process of this embodiment is as follows: initialize parameters, initialize the initial value of the population according to the exponential distribution, generate the offspring population according to formula (15), use the information feedback mechanism to update the individual value until all individuals in the population are updated, merge the parent generation and the offspring, perform non-dominated sorting and divide the non-dominated level, select individuals with high dominance level, complete the population number according to the crowding sorting, check whether the number of iterations has been reached, if it has been reached, end the iteration and output the result, if not, continue the iteration.

[0033] Embodiment 3: The pavement performance prediction module of this embodiment adopts a physical embedded neural network. Figure 2It is a schematic diagram of the physical embedded neural network structure, which includes four main parts: input layer, hidden layer, output layer and loss term. Among them, the input layer includes road age, traffic factors (average annual passenger car traffic volume, passenger-cargo ratio, average annual truck traffic volume, equivalent load times), geological factors (base properties, geological structure, soil characteristics), weather and environmental factors (rainfall, annual average daily maximum temperature, annual average daily minimum temperature), road structure (surface material, surface thickness, base material, base thickness), maintenance methods (micro-surfacing, pavement patching, thermal regeneration, single-layer excavation and patching, single-layer overlay, double-layer excavation and patching, double-layer overlay, three-layer excavation and patching, roadbed grouting) and pavement performance values ​​in the past year, and the output value is the pavement performance in the next year. The input vector of each layer in the physical embedded neural network comes from the output vector of the previous layer, and the input vector is calculated by weight and bias processing to form the output vector.

[0034] The loss item It consists of four parts, and the calculation formula is: (19) in, Represents all parameters of the physical embedded neural network, is the data loss term coefficient, is the data loss term, is the physical loss term coefficient, is the physical loss term, is the boundary loss term coefficient, is the boundary loss term, is the regular loss term coefficient, is the regular loss term.

[0035] The data loss term is the error between the predicted value and the true value and is calculated as follows: (20) in, represents the physical embedded neural network prediction value of the sth sample, represents the true value of the data item of the sth sample, represents the number of samples used to calculate the data item, Input vector representing the sth sample data item.

[0036] The physical loss term is calculated according to the following formula: (twenty one) in, represents the predicted value of the physical item of the sth sample, represents the true value of the physical item of the sth sample, represents the number of samples used to calculate the physical term, Input vector representing the sth sample physical term. The calculation formula is as follows: (twenty two) in , ,…, is a parameter to be determined, t represents time, is the input variable, n represents the number of input variables; PPI is the pavement performance, Represents the physical embedded neural network as a whole.

[0037] The calculation formula of the regularity loss term is as follows: (twenty three) in, represents the number of samples used to calculate the regularity loss, represents the input vector of the s-th sample regularity term, Represents the partial derivative of the physical embedded neural network of the sth sample with respect to time t. The regular meaning of this regular loss term is that the road performance decreases over time.

[0038] The boundary loss term is calculated as follows: (twenty four) in, represents the number of samples used to calculate the boundary loss, Represents the output value of the physical embedded neural network when the input vector is a 0 vector, Represents the partial derivative of the physical embedded neural network with respect to time t when the input vector is 0.

[0039] Embodiment 4: This embodiment also includes a data preprocessing module, and the data preprocessing process is as follows: (1) The input data is processed in time series. Each pile number and its corresponding data are arranged vertically according to time. The influencing factor data of each pile number are arranged horizontally. The last column is the pavement performance value.

[0040] (2) Process the data in a supervised manner and arrange the next row of data in each row horizontally behind the row, so that each row contains the influencing factors and pavement performance data of the pile number in the current year and the influencing factors and pavement performance data of the next year.

[0041] (3) Clean up abnormal data, align the pile number, year, and lane of each row of data, and remove data with a maintenance factor of null but the pavement performance in the next year will be higher than that in the current year, and data with a maintenance factor of not null but the pavement performance in the next year will be lower than that in the current year.

[0042] (4) Use one-hot encoding to convert the maintenance method data into a vector of 0 and 1. The nine maintenance methods are encoded as follows: micro-surfacing [1,0,0,0,0,0,0,0,0], pavement lining [0,1,0,0,0,0,0,0,0], thermal regeneration [0,0,1,0,0,0,0,0,0], single-layer excavation and patching [0,0,0,1,0,0,0,0,0], single-layer overlay [0,0,0,0,1,0,0,0,0], double-layer excavation and patching [0,0,0,0,0,1,0,0,0], double-layer overlay [0,0,0,0,0,0,1,0,0], triple-layer excavation and patching [0,0,0,0,0,0,0,1,0], and roadbed grouting [0,0,0,0,0,0,0,0,1]. The subsequent maintenance plan matrix is ​​also processed according to this coding method.

[0043] (5) Divide the data by time and normalize the data. Divide the data by time, with the first N-1 years of data as the training set data and the last year of data as the test set data. The purpose of data normalization is to map all data to between 0 and 1 so that the neural network can learn the data features. The specific data normalization formula is as follows: (25) in, is the maximum value of the current column. is the minimum value of the current column, and X is the data of the current column; is the value after data normalization.

[0044] After the above data preprocessing, a data set is formed, and the data set is used to train the physical embedded neural network architecture of the road performance prediction module. The neural network training process is: (1) This embodiment adopts the physical embedded neural network of Embodiment 3 and establishes the physical embedded neural network architecture using Python language.

[0045] (2) Set the hyperparameters of the physical embedded neural network. Set the hidden layer to three fully connected layers, set the number of physical embedded neural network units in each hidden layer to 64, set the activation function to the tanh function, set the batch size to 128, set the learning rate to 0.001, set the number of training times to 2000, set the Adam automatic differentiation calculation, and set the loss function to formula (19).

[0046] (3) Set the last column of the data as the output value and the other columns as the input values, and call the fit function of the keras library for training.

[0047] (4) Call the save function in the keras library to save the model parameters.

[0048] When predicting, ensure that the input data composition is consistent with the training data, and call the load_model function in the keras library to make predictions.

[0049] After the training is completed, a trained model is obtained, and the trained model is used for pavement performance prediction, and the predicted value is used as the basis of the pile number level fitness calculation module.

[0050] Embodiment 5: In this embodiment, the structure of the road section level fitness calculation module is similar to that of the road network level fitness calculation module, and the road section level fitness calculation module is taken as an example for explanation: In the road section-level fitness calculation module, the calculation formulas for the cost CL, carbon emission CEL, comprehensive pavement performance CRPSL and construction time CTL of each road section are as follows: (26); (27) (28) (29) Where n represents the number of pile numbers; h represents the hth 100-meter pile number in the road section; represents an optimal maintenance plan in the Pareto optimal maintenance plan set for the pile number h; It represents an optimal maintenance plan in the Pareto optimal maintenance plan set for the pile number h+1; , The optimal maintenance plan for the section from pile number h to pile number h+1 is arrive Additional costs and construction time; , , , They represent the optimal maintenance plan M of the pile number h in the Pareto optimal maintenance plan set. h Cost, carbon emissions, pavement performance, and construction time.

[0051] In the road network level fitness calculation module, n represents the number of road sections; h represents the hth road section in the road network; An optimal maintenance plan in the Pareto optimal maintenance plan set for the h-th road section; represents an optimal maintenance plan in the Pareto optimal maintenance plan set for the h+1th road section; , They represent the optimal maintenance plan for switching from section h to section h+1. arrive Additional costs and construction time; , , , They represent the optimal maintenance plan M of the Pareto optimal maintenance plan set for road section h. h Cost, carbon emissions, pavement performance, and construction time.

[0052] In the pile number level fitness calculation module, the maintenance specification recommends that: when the predicted pavement performance value is below 90, micro-surfacing is not recommended, and when the predicted pavement performance value is below 80, covering measures (single-layer covering or double-layer covering in the present invention) are not recommended. The maintenance plan is composed of a maintenance plan matrix according to the aforementioned unique hot code, and the elements in the matrix are represented in sequence as follows: micro-surfacing, pavement lining, thermal regeneration, single-layer excavation and patching, single-layer covering, double-layer excavation and patching, double-layer covering, three-layer excavation and patching, and roadbed grouting. Therefore, this embodiment adopts the following restrictions in the pile number level fitness calculation module according to the maintenance specifications: (30) in Indicates that the micro-surfacing (first type) maintenance method is not adopted in year t. Similarly, Indicates that the single-layer covering (fourth type) maintenance method is not adopted in year t. It means that the double-layer covering (seventh type) maintenance method will not be adopted in year t.

[0053] Similarly, in the road segment-level fitness calculation module and the road network-level fitness calculation module, the comprehensive pavement performance of each pile number and the comprehensive pavement performance of each road segment also meet the restrictions recommended by the maintenance specifications.

[0054] In addition, the present invention stipulates that the cost, carbon emissions, and construction time of the maintenance plan have upper limits, and the comprehensive pavement performance has lower limits, as shown in the following formula: (31) in, represents the cost cap, Represents the upper limit of construction time, represents the carbon emission cap, Represents the lower limit of comprehensive performance.

[0055] When the optimal maintenance plan provided by the optimization module does not meet the above restrictions, the fitness will provide an infinite value to indicate that this type of maintenance plan is unreasonable.

[0056] Embodiment 6: The specific process of the physical data dual-driven road maintenance decision system in this embodiment is as follows: Figure 3 As shown: Step 1. Preprocess the data and train the physical embedded neural network to obtain the pavement performance prediction module.

[0057] Step 2. Enter the pre-maintenance pile number, road section and road network information. Enter the pre-maintenance pile number, road section and road network lane pile number position pre-maintenance year, and pavement performance data in pile numbers.

[0058] Step 3. Pile maintenance decision stage: The maintenance plan of each pile number is optimized (the maintenance plan is what kind of maintenance method is adopted each year in N years. The maintenance plan matrix is ​​the digital expression of the maintenance plan. If there is no maintenance in the first year, the first row is all 0. If the first maintenance method is adopted in the first year, the first column of the first row is 1, and the other columns are 0. And so on, a maintenance plan matrix consisting of 0 and 1 with N rows and M columns is formed). After optimizing the maintenance plan, the pile number level optimization module inputs the maintenance plan into the pavement performance prediction module and the pile number level fitness calculation module respectively. The pavement performance prediction module will output the pavement performance prediction value based on the influencing factor data matching the maintenance section location and year information.

[0059] The pile number level fitness calculation module calculates the cost, carbon emission, construction time, and comprehensive pavement performance of the pile number under the maintenance plan according to the maintenance plan provided by the pile number level optimization module and the pavement performance prediction value provided by the pavement performance prediction module, and uses this as the pile number level fitness and returns it to the pile number level optimization module. The pile number level optimization module then sorts the population by non-dominated order and congestion order according to the pile number level fitness, selects a new generation of individuals (each individual corresponds to a maintenance plan), and provides the maintenance plan of the new generation of individuals to the pile number level fitness calculation module and the pavement performance prediction module. Repeat the above process until the iteration stop condition is met, and output the pile number Pareto optimal maintenance plan set for each pile number. And the pile number Pareto optimal maintenance plan set for each pile number is input into the section level optimization module in the section maintenance decision stage.

[0060] Step 4. Section maintenance decision stage: Optimize the maintenance plan for each road section. The section-level optimization module takes the Pareto optimal maintenance plan set of each pile number in the entire road section as the population individual, and provides the maintenance plan of the road section to the section fitness calculation module after initialization. The section-level fitness calculation module calculates the cost, carbon emission, construction time, and comprehensive pavement performance of the road section according to the maintenance plan of the road section, and uses this as the section-level fitness, and returns it to the section-level optimization module. The section-level optimization module sorts the population by non-dominated and congestion according to the section-level fitness, selects a new generation of individuals (each individual corresponds to a maintenance plan), and provides the maintenance plan of the new generation of individuals to the section-level fitness calculation module. Repeat the above process until the iteration stop condition is met, and output the section Pareto optimal maintenance plan set for each road section. And the section Pareto optimal maintenance plan set for each road section is input into the road network-level optimization module in the road network maintenance decision stage.

[0061] Step 5. Road network maintenance decision-making stage: Optimize the maintenance plan of the entire road network. The road network optimization module takes the Pareto optimal maintenance plan set of each road section in the entire road network as the population individual, initializes the maintenance plan of the road network and provides it to the road network fitness calculation module. The road network fitness calculation module calculates the cost, carbon emissions, construction time, and comprehensive pavement performance of the road network according to the maintenance plan of the road network, and uses this as the road network fitness and returns it to the road network optimization module. The road network optimization module sorts the population by non-dominated and congestion according to the road network fitness, selects a new generation of individuals (each individual corresponds to a maintenance plan), and provides the maintenance plan of the new generation of individuals to the road network fitness calculation module. Repeat the above process until the iteration stop condition is met, and output the Pareto optimal maintenance plan set of the road network.

[0062] The above iteration stopping condition may be whether a set maximum number of iterations is reached.

[0063] Any matters not described in the present invention are applicable to the prior art.

Claims

1. A road maintenance decision system driven by both physical data and physical data, characterized in that: The system comprises: The pavement performance prediction module is used to predict the pavement performance in the next year based on the road age, traffic factors, geological factors, weather and environmental factors, road structure, maintenance methods and the pavement performance values ​​in the past year; The pile number level fitness calculation module is used to calculate the cost, carbon emission, construction time and comprehensive pavement performance of each pile number according to the maintenance plan output by the pile number level optimization module and the pavement performance predicted by the pavement performance prediction module as the pile number level fitness; The pile number optimization module uses a multi-objective exponential distribution optimization algorithm to perform multi-objective optimization on the pile number fitness calculated by the pile number fitness calculation module, outputs a maintenance plan, and obtains the pile number Pareto optimal maintenance plan set for each pile number after reaching the number of iterations. The pile number Pareto optimal maintenance plan set includes G optimal maintenance plans; A section-level fitness calculation module, used to calculate the cost, carbon emission, construction time and comprehensive pavement performance of each section according to the Pareto optimal maintenance plan set of the pile number and the comprehensive pavement performance of each pile number obtained by the pile number-level fitness calculation module, as the section-level fitness; The section-level optimization module inputs the Pareto optimal maintenance plan set of each pile number, and uses the multi-objective exponential distribution optimization algorithm to perform multi-objective optimization on the section-level fitness calculated by the section-level fitness calculation module, outputs the maintenance plan, and obtains the Pareto optimal maintenance plan set of each section after reaching the number of iterations; A road network-level fitness calculation module, used to calculate the cost, carbon emission, construction time and comprehensive road performance of the road network according to the Pareto optimal maintenance plan set of the road section and the comprehensive road performance of each road section obtained by the road section-level fitness calculation module, as the road network-level fitness; The road network level optimization module takes the Pareto optimal maintenance plan set of the road section as input, uses the multi-objective exponential distribution optimization algorithm to perform multi-objective optimization on the road network level fitness calculated by the road network level fitness calculation module, outputs the maintenance plan, and obtains the road network Pareto optimal maintenance plan set after reaching the number of iterations.

2. The physical data dual-driven road maintenance decision system according to claim 1 is characterized in that: The pavement performance prediction module uses a physical embedded neural network PINN for prediction, including an input layer, a hidden layer, an output layer and a loss term; The loss item It consists of four parts, and the calculation formula is: （19） in, Represents all parameters of the physical embedded neural network, is the data loss term coefficient, is the data loss term, is the physical loss term coefficient, is the physical loss term, is the boundary loss term coefficient, is the boundary loss term, is the regular loss term coefficient, is the regular loss term.

3. The physical data dual-driven road maintenance decision system according to claim 1 is characterized in that: In the pile number level fitness calculation module, the calculation formulas for the cost Cost, carbon emission CE, construction time CT and comprehensive pavement performance CRPS of each pile number are as follows: （4） （5） （6） （7） in, It represents the cost of each maintenance plan corresponding to each pile number to adopt the g-th maintenance method in the tth year, t0 represents the starting year, t e Indicates the year of the end; It represents the carbon emission of each maintenance scheme corresponding to each pile number when the g-th maintenance method is adopted in the t-th year; It indicates the time required for each maintenance scheme corresponding to each pile number to be constructed by the g-th maintenance method in the t-th year; the value of g is an integer from 1 to 9, representing micro-surfacing, pavement lining, thermal regeneration, single-layer excavation and patching, single-layer overlay, double-layer excavation and patching, double-layer overlay, triple-layer excavation and patching, and roadbed grouting respectively; Represents the physical embedded neural network in the input vector The predicted value below.

4. The physical data dual-driven road maintenance decision system according to claim 1 is characterized in that: In the road section-level fitness calculation module, the calculation formulas for the cost CL, carbon emission CEL, comprehensive pavement performance CRPSL and construction time CTL of each road section are as follows: （26） （27） （28） （29） Where n represents the number of pile numbers; h represents the hth 100-meter pile number in the road section; represents an optimal maintenance plan in the Pareto optimal maintenance plan set for the pile number h; It represents an optimal maintenance plan in the Pareto optimal maintenance plan set for the pile number h+1; , The optimal maintenance plan for the section from pile number h to pile number h+1 is arrive Additional costs and construction time; , , , They represent the optimal maintenance plan M of the pile number h in the Pareto optimal maintenance plan set. h Cost, carbon emissions, pavement performance, and construction time.

5. The physical data dual-driven road maintenance decision system according to claim 1 is characterized in that: The traffic factors include the average annual passenger car traffic volume, passenger-to-freight ratio, average annual truck traffic volume and equivalent load times; the geological factors include base properties, geological structure and soil characteristics; the weather environment factors include rainfall, annual average daily maximum temperature and annual average daily minimum temperature; the road structure includes surface layer materials, surface layer thickness, base layer materials and base layer thickness; the maintenance methods include micro-surfacing, pavement lining, thermal regeneration, single-layer excavation and patching, single-layer overlay, double-layer excavation and patching, double-layer overlay, triple-layer excavation and patching and roadbed grouting.

6. The physical data dual-driven road maintenance decision system according to claim 1 is characterized in that: The multi-objective exponential distribution optimization algorithm in the pile level optimization module includes the following process: initializing a population P, the number of individuals in the population is K, and each individual is an N×M maintenance plan matrix, where N is the length of the maintenance year, M is the number of maintenance methods, and each individual in the population P Represents the N-year maintenance plan of the pile number, where each individual maintenance plan satisfies that only one maintenance method is adopted each year, and each row of the maintenance plan matrix satisfies the constraint shown in formula (9): （9） Initialize the individual according to formula (10): （10） Among them, rand(1,M) represents randomly selecting an index column from {1,2,…,M} according to exponential distribution; The exploration and exploitation principle is used to calculate the guided solution according to formula (11). : （11） Among them, time represents the number of iterations, , , Represents the three best solutions at the current number of iterations; The output results of the pile level fitness calculation module are used for information feedback. The update logic of each individual maintenance plan is as follows: the weighted coefficients of the guided solution and the memoryless solution are obtained based on the pile level fitness to calculate the value of the new individual matrix. , and The maximum value of each row is assigned to 1, and the other values ​​are assigned to 0. Then the next generation of individuals is updated according to formula (12): （12） in, It represents selecting the index column j with the largest value for the i-th row of the current matrix, and the corresponding position of the index column with the largest value is assigned 1; is the updated individual of the i-th row, j-th column and time+1-th iteration; is defined as: （13） in, is the memoryless solution of the i-th row and j-th column in the time-th generation individual; a and b are weighted coefficients, which are defined as: （14） in, is the fitness of the guiding solution, Represents the fitness of the kth individual, and both are obtained by the pile-level fitness calculation module.

7. The physical data dual-driven road maintenance decision system according to claim 1 is characterized in that: The pile number level fitness calculation module, section level fitness calculation module, and road network level fitness calculation module all use the maintenance specification recommendations as restriction conditions. When the pavement performance value is below 90, micro-surfacing will not be used, and when the pavement performance value is below 80, covering measures will not be used.

Images