County and city built-up area road traffic energy consumption prediction method

Abstract

The invention provides a county built-up area road traffic energy consumption prediction method, which comprehensively considers urban morphological characteristics, and comprises the following steps: step 1, constructing a road traffic energy consumption simulation model; and 2, constructing a road traffic energy consumption prediction model to predict the road traffic energy consumption. By applying the technical scheme, the accuracy and practicability of prediction can be improved, and urban and rural planning and traffic management decisions can be better served.

Description

Technical Field

[0001] The invention relates to the technical field of urban construction, in particular to a method for predicting road traffic energy consumption in a county town built-up area. Background Art

[0002] Current urban technology research is mainly concentrated on two levels: macro and micro. Research at the macro level attempts to predict the overall traffic energy consumption of the city by comprehensively considering a variety of relevant factors, but the current prediction accuracy needs to be improved, and it is difficult to accurately reflect the actual energy consumption status of the urban space. At the micro level, research usually focuses on factors such as road conditions, vehicle speed, vehicle performance, and driving conditions, using traffic simulation technology to evaluate the fuel consumption of specific vehicles or modes of transportation per unit time. However, these studies often lack a close connection with the specific spatial characteristics of the city. Whether it is a macro or micro level traffic energy consumption model, currently there is little consideration of urban block-level planning elements, which limits the application potential of the model in urban and rural planning practice. Summary of the invention

[0003] In view of this, the purpose of the present invention is to provide a method for predicting road traffic energy consumption in built-up areas of county towns, so as to improve the accuracy and practicality of the prediction and better serve urban and rural planning and traffic management decisions.

[0004] To achieve the above purpose, the present invention adopts the following technical solution: a method for predicting road traffic energy consumption in a county built-up area, comprising the following steps:

[0005] Step 1: Construct a road traffic energy consumption simulation model;

[0006] Step 2: Construct a road traffic energy consumption prediction model to predict road traffic energy consumption.

[0007] In a preferred embodiment, the step 1 specifically includes:

[0008] Step 11: Divide the traffic energy consumption analysis area; divide the county built-up area into several sub-areas; when dividing, the overall block should be divided according to the road shape;

[0009] Step 12: Determine the county center area;

[0010] Step 13: Determine the urban morphology variable indicators; select the urban morphology indicators population density and POI facilities that have the greatest impact on urban road traffic volume as the urban morphology input variables of the traffic energy consumption simulation model;

[0011] Step 14: Calculate location advantage;

[0012] Step 15: Calculate traffic energy consumption;

[0013] Step 16: Calculation of travel weight POI;

[0014] Step 17: Analyze the distribution of traffic energy consumption in the area;

[0015] Step 18: After calculating the energy consumption scores of each zone, the values ​​are distributed to the road network to obtain the traffic energy consumption value of each road section;

[0016] Step 19: Use field measurements of road traffic volume to verify the accuracy of the model.

[0017] In a preferred embodiment, step 14 specifically includes: introducing a plot accessibility correction index to improve the traffic generation calculation of the plot unit; defining the plot unit accessibility to be determined by the planned road network density around the plot and the spatial distance between the plot and the central node within the planning range, as shown in the following formula:

[0018]

[0019] Where: K i is the accessibility of the analysis area i; P i P is the road network density around the analysis area i. According to the scale of the analysis area, the surrounding analysis area is appropriately selected to calculate its road network density; max is the road network density value of the analysis area with the maximum road network density within the research scope; L i is the spatial distance between the centroid of the analysis area i and the center point, which has been obtained through the standard ellipse difference analysis of the POI point in the above text; L min It is the minimum value of the distance between all analysis areas and the center point within the research scope.

[0020] In a preferred embodiment, the step 15 specifically includes: designing a travel volume prediction model based on location advantage, population density, number of parking spaces, and POI facilities as follows:

[0021]

[0022] Among them, A i is the traffic volume of analysis area i; m represents the number of POI categories except residential and company categories; a q Represents the travel weight of the POI point of category q; S q represents the number of POI points of category q in the analysis area i; i represents the POI travel generation rate of the analysis area i; L represents the number of permanent residents in the study area; n represents the number of residential communities in the analysis area i; a r represents the travel incidence rate based on residential areas; L r represents the number of residents in the rth residential area in the analysis area; a j represents the travel incidence rate based on office companies; L jrepresents the number of employees in the jth company in the analysis area; K i is the locational advantage of analysis area i;

[0023] Since the amount of traffic energy consumption generated is positively correlated with the travel volume, the traffic energy consumption attraction weight of the analysis area is expressed as follows:

[0024]

[0025] Among them, T i is the traffic energy consumption attraction weight of analysis area i, A i is the traffic volume of analysis area i, n is the number of analysis areas;

[0026] Using refined oil sales data as the total value of the county road network traffic energy consumption, the traffic energy consumption generated by the sub-area is as follows:

[0027] P i =P×T i (4)

[0028] Among them, P i is the traffic energy consumption generated in the analysis area i, P is the total sales volume of refined oil, T i is the traffic energy consumption attraction weight of analysis area i.

[0029] In a preferred embodiment, the step 16 specifically includes: the ratio of the POI facility point in each analysis area to the traffic volume in the built-up area is defined as the POI travel generation rate, which can be expressed as the following formula:

[0030]

[0031] Among them, O i represents the POI travel generation rate of the analysis area i; m represents the number of POI categories except residential and company categories; a q Represents the travel weight of the POI point of category q; S q represents the number of POI points of category q in analysis area i; n represents the number of analysis areas.

[0032] In a preferred embodiment, the step 2 specifically includes:

[0033] Step 21: Determine the input variables and output variables of the neural network;

[0034] Step 22: Select the variables of the prediction model. The input layer includes three variables: the number of POI points, the location advantage value, and the road grade.

[0035] Step 23: Before entering the variables, outliers need to be processed to ensure that the data remains within the appropriate range;

[0036] Step 24: Select a three-layer BP neural network to meet the model requirements, and select 1 as the number of hidden layers;

[0037] Step 25: Select activation function;

[0038] Step 26: Select the steepest descent BP algorithm to optimize the model.

[0039] In a preferred embodiment, the step 21 specifically includes: the algorithm formula of the BP neural network involves two processes: forward propagation and back propagation.

[0040] In a preferred embodiment, during the forward propagation process, data goes from the input layer through the hidden layer to the output layer; let the weight between node i and node i be w ij , the threshold of the node is b j , the output value of each node is x j , then for each neuron, its output is calculated by the following formula:

[0041]

[0042] x j =f(S j ) (6)

[0043] Among them, f is the activation function.

[0044] In a preferred embodiment, the back-propagation process is based on the gradient descent method, which updates the weights and biases of the network by back-propagating the error layer by layer, thereby gradually reducing the prediction error; in this process, the error signal starts from the output layer and propagates back to the input layer layer by layer, and the weights and biases in the network are adjusted according to these error signals.

[0045] In a preferred embodiment, a loss function E is defined to quantify the difference between the network prediction output and the actual output; the mean square error MSE is a common form of the loss function, which is a numerical indicator reflecting the prediction accuracy by calculating the square of the difference between the predicted value and the actual value and finding the average, which can effectively measure the degree of deviation between the predicted value and the actual value; for a single sample, the mean square error loss function can be specifically defined as:

[0046]

[0047] Where M is the number of neurons in the output layer; is the actual output of the jth neuron in the network output layer; y j is the corresponding target output;

[0048] Calculate the partial derivatives of the loss function with respect to the weights and biases, i.e. the gradients; for the weights and biases of the output layer, a specific algorithm is used to calculate these gradient values:

[0049]

[0050] Where L represents the number of output layers;

[0051]

[0052] is the weighted input of the jth neuron in the output layer;

[0053] is the output of the jth neuron in the output layer, where f is the activation function;

[0054] The update formula for weights and biases is expressed as:

[0055]

[0056] in, is the updated value of weight w; is the updated value of bias b; η is the learning rate.

[0057] Compared with the prior art, the present invention has the following beneficial effects: the road traffic energy consumption model constructed by the present invention can facilitate researchers in urban planning and related fields to study urban traffic energy consumption at the scale of urban blocks, and is also convenient for optimizing the layout of urban traffic. The significant advantages of this model include: 1) Intuitive display: By combining urban morphological elements, the model can intuitively display the distribution of traffic energy consumption in different regions, allowing researchers and decision makers to quickly identify energy consumption hotspots and potential optimization areas. 2) Rapid prediction: Using advanced BP neural network technology, the model can quickly predict the energy consumption value of a specific location by only inputting relevant values ​​of urban morphology, greatly improving the efficiency of research and planning. 3) Layout optimization: The output results of the model help guide the optimization of urban traffic layout, and support more scientific and sustainable urban development strategies by accurately predicting energy consumption under different planning schemes. 4) Decision support: Provide data-based decision support for policy makers and urban planners to help them formulate effective energy-saving and emission reduction measures and promote the green transformation of urban transportation systems. 5) Research deepening: The application of the model not only promotes a deeper understanding of the factors affecting urban traffic energy consumption, but also provides a new perspective for exploring the complexity of the relationship between urban spatial form and energy consumption. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] Figure 1 A flow chart of a road traffic energy consumption simulation model according to a preferred embodiment of the present invention;

[0059] Figure 2A flow chart of a neural network prediction model for road traffic energy consumption according to a preferred embodiment of the present invention;

[0060] Figure 3 It is an overall flow chart of a preferred embodiment of the present invention. DETAILED DESCRIPTION

[0061] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0062] It should be noted that the following detailed descriptions are illustrative and are intended to provide further explanation of the present application. Unless otherwise specified, all technical and scientific terms used in the present invention have the same meanings as those commonly understood by those skilled in the art to which the present application belongs.

[0063] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present application; as used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprise" and / or "include" are used in this specification, they indicate the presence of features, steps, operations, devices, components and / or their combinations.

[0064] refer to Figure 1-3 The invention aims to construct a road traffic energy consumption prediction model combined with urban morphology, use the refined oil sales data of gas stations and urban morphology data to predict road traffic energy consumption, and finally form a road traffic energy consumption prediction model including the urban road network structure. The road traffic energy consumption model proposed in this study consists of two parts: a road traffic energy consumption simulation model and a road traffic energy consumption prediction model. These two models use the gravity model and BP neural network as the core methods respectively.

[0065] The following are the steps to build this model:

[0066] 1. Steps for building a road traffic energy consumption simulation model:

[0067] The road traffic energy consumption simulation model constructed by the present invention is constructed in the following steps: Figure 1 , the specific steps are as follows:

[0068] 1.1 Divide the traffic energy consumption analysis area:

[0069] Refer to the division method of traffic zones to ensure that the social, economic and land use conditions in the analysis area are as consistent as possible, and fully consider the layout of the road network, the integrity of the zones, and avoid dividing the land with the same use. The built-up area of ​​the county town is divided into several zones. When dividing, the overall block should be divided according to the shape of the road. Based on the traffic volume in different areas, the zones with large traffic volume have smaller areas, and the zones with small traffic volume have larger areas. The zones near the city center have smaller areas, and the zones near the suburbs have larger areas, so as to ensure the simulation accuracy of traffic energy consumption.

[0070] 1.2 Determine the county center area:

[0071] By identifying the center point through ellipse difference analysis of urban POI facility points, the spatial distribution of different POI types can be intuitively displayed, and their distribution characteristics can be analyzed to obtain the core area and center point of public activities (i.e., human flow and traffic activities) in the county.

[0072] 1.3 Determine the urban morphology variable indicators:

[0073] The present invention selects population density and POI facilities, the urban morphology indicators that have the greatest impact on urban road traffic volume, as urban morphology input variables of the traffic energy consumption simulation model. Population is the decisive factor in traffic generation, while POI facilities can characterize the nature, density and other characteristics of land use. They are comprehensive and easy to obtain, and are important indicators for studying urban morphology. Combined with the characteristics of the built-up area of ​​the county, traffic travel can be divided into commuting travel based on residences and companies and non-commuting travel based on other POI facilities such as catering, leisure, shopping, culture and entertainment. The population density variable can be further divided into two variables: residential population density and office population density. Based on this idea, the urban morphology indicators selected in the subsequent road traffic energy consumption simulation model are shown in Tables 1 and 2.

[0074] Table 1 Urban morphological variables used in the traffic energy consumption simulation model

[0075]

[0076] Table 2 POI classification table

[0077]

[0078] 1.4 Calculate location advantage:

[0079] The present invention comprehensively considers the two factors of location and accessibility, and introduces a plot accessibility correction index to improve the calculation of traffic generation of plot units. The present invention defines that the plot unit accessibility is determined by the planned road network density around the plot and the spatial distance between the plot and the central node within the planning range, as shown in the following formula:

[0080]

[0081] Where: K i is the accessibility of the analysis area i; P i is the road network density around the analysis area i. The road network density of the surrounding analysis area can be calculated by appropriately selecting the surrounding analysis area according to the scale of the analysis area; max is the road network density value of the analysis area with the maximum road network density within the research scope; L i is the spatial distance between the centroid of the analysis area i and the center point, which has been obtained through the standard ellipse difference analysis of the POI point in the above text; L min It is the minimum value of the distance between all analysis areas and the center point within the research scope.

[0082] 1.5 Calculation of traffic energy consumption:

[0083] The travel volume prediction model designed based on location advantage, population density, number of parking spaces, POI facilities, etc. is as follows

[0084]

[0085] Among them, A i is the traffic volume of analysis area i; m represents the number of POI categories except residential and company categories; a q Represents the travel weight of the POI point of category q; S q represents the number of POI points of category q in the analysis area i; i represents the POI travel generation rate of the analysis area i; L represents the number of permanent residents in the study area; n represents the number of residential communities in the analysis area i; a r represents the travel incidence rate based on residential areas; L r represents the number of residents in the rth residential area in the analysis area; a j represents the travel incidence rate based on office companies; L j represents the number of employees in the jth company in the analysis area; K i is the location advantage of analysis area i.

[0086] Since the amount of traffic energy consumption generated is positively correlated with the travel volume, the traffic energy consumption attraction weight of the analysis area can be expressed as follows:

[0087]

[0088] Among them, T i is the traffic energy consumption attraction weight of analysis area i, A i is the traffic volume of analysis area i, and n is the number of analysis areas.

[0089] Since the present invention uses the refined oil sales data as the total value of the county road network traffic energy consumption, the traffic energy consumption generated by the sub-area is as follows:

[0090]

[0091] Among them, Pi is the traffic energy consumption generation in analysis area i, P is the total sales volume of refined oil, and Ti is the traffic energy consumption attraction weight in analysis area i.

[0092] 1.6 POI travel weight:

[0093] The average number of daily traffic trips of residents based on residential areas is 1.66, so the trip occurrence rate based on residential areas is 1.66 / person. According to existing research, the trip attraction rate based on companies can be taken as 0.96 person-times / employee. When calculating the trip occurrence of other POI facilities, given the differences in importance of different POI types, different weights need to be assigned to each type of POI when conducting a comprehensive analysis.

[0094] The proportion of traffic volume attracted by POI facilities in each analysis area in the built-up area is defined as the POI travel generation rate, which can be expressed as follows:

[0095]

[0096] Among them, O i represents the POI travel generation rate of the analysis area i; m represents the number of POI categories except residential and company categories; a q Represents the travel weight of the POI point of category q; S q represents the number of POI points of category q in analysis area i; n represents the number of analysis areas.

[0097] Table 3 POI travel weight assignment

[0098]

[0099] 1.7 Analysis of regional traffic energy consumption distribution:

[0100] Substituting the relevant data of the county area into equations (1), (2), (3), (4) and (5), we can obtain the traffic energy consumption allocation value of the county area.

[0101] 1.8 Traffic energy consumption distribution based on gravity model:

[0102] After calculating the energy consumption scores of each partition, the values ​​need to be allocated to the road network to obtain the traffic energy consumption value of each road section. This is mainly done using the gravity model (Huff model) in the location allocation model in the GIS software. The purpose of the present invention is to use the location allocation method to allocate the known total traffic energy consumption to each road facility and analyze the distribution differences, which is similar to the competitive relationship between candidate points in general analysis. Therefore, the allocation method adopted is the maximization market share analysis method.

[0103] First, you need to create a network dataset for the road network in the study area in GIS. The steps are as follows: Step 1, collect data such as the location, length, and speed limit of the road on the OpenStreetMap map website. Step 2, use the Network Dataset Wizard tool in ArcGIS software to build a network dataset based on the collected road network data. Step 3, define network attributes, set speed limits, driving directions, road grades, etc. Step 4, use the Topology Builder tool in ArcGIS software to establish the topological relationship of the network dataset and check its connectivity. Step 5, verify and test the network dataset through the Network Analyst tool in ArcGIS software to ensure its availability and accuracy.

[0104] Then, in the GIS software, use the location allocation tool (Location Allocation) to create a new location allocation model. Load the demand points and facility points. The road partition centroid is used as the demand point, and the road point feature is used as the facility point. The demand point allocation weight is P i The impedance parameter is set to 2, which means that P i It will decay with the square of the road distance. Since the present invention does not screen the facility points, but requires the energy consumption allocation value of each facility point, the impedance interruption coefficient can be set to infinity. According to the driving habits of cars, the attraction of the main road to cars should be greater than that of the secondary road, followed by the branch road. Therefore, the facility point allocation weight is divided into 0.5, 1 and 2 according to the road grade, corresponding to the branch road, secondary road and main road respectively. Running the current analysis, the traffic energy consumption generated by the demand point P can be obtained i The distribution value at each facility in the road network is displayed as the DemandWeight field.

[0105] Finally, the road facility point data is connected to the road segments through the inter-connection function of GIS. The traffic energy consumption summary value of the road segment is visualized using the vector data visualization function of GIS. The final road network traffic energy consumption distribution map is obtained.

[0106] 1.9 Model verification and correction:

[0107] In order to verify the accuracy of the model, it is necessary to obtain the traffic energy consumption value of the road section based on the actual traffic volume of the road. At present, the main idea for calculating the traffic energy consumption of road networks at home and abroad is the product of the fuel consumption rate and emission rate and the total mileage of the road. Combined with the literature, the following method can be used to calculate the traffic energy consumption of different specific sections, and the traffic energy consumption model per unit time of class z vehicles in section j can be approximately obtained to verify the accuracy of the model:

[0108] P zj =q zj × j ×Ef zj (6)

[0109] Where: P zj is the traffic energy consumption of class z vehicles on road section (kg); q zj is the traffic volume of category z on the road section (pcu / h); l j is the length of the road section (km); E fzj is the average fuel consumption of class z vehicles on the road section (L / km). Taking passenger cars as the standard, the average fuel consumption of vehicles can be taken as 0.8L / KM.

[0110] At the same time, field measurements of road traffic volume are needed to verify the accuracy of the model.

[0111] 2. Steps for building a road traffic energy consumption prediction model:

[0112] The present invention simulates and calculates the current road traffic energy consumption. In order to further optimize the calculation of the model, the energy consumption value of a certain location can be quickly obtained by inputting relevant values ​​of certain urban forms. This can help quickly understand the traffic energy consumption of the urban road network during urban planning, so as to make better traffic planning and policy formulation. Here, this study introduces the BP neural network, that is, the back propagation neural network model, to further predict and optimize the road network energy consumption model.

[0113] The present invention uses MATLAB2022a software to construct a BP neural network. MATLAB2022a is an online data processing and analysis platform that integrates professional statistical methods and data algorithms. It can be widely used in scientific research, data mining and other fields. It is a powerful tool for building neural network models. It supports the setting of hyperparameters such as learning rate, number of hidden layers, number of neurons in each hidden layer, and selection of activation functions, and can easily and quickly obtain the calculation results of the BP neural network.

[0114] The road traffic energy consumption simulation model constructed by the present invention is constructed in the following steps: Figure 2, the specific steps are as follows: 2.1 BP neural network algorithm introduction:

[0115] The algorithm formula of BP neural network mainly involves two processes: forward propagation and back propagation. The following are the main formulas of these two processes:

[0116] In the forward propagation process, data goes from the input layer through the hidden layer to the output layer. Let the weight between node i and node i be wij, the threshold of the node be bj, and the output value of each node be xj. For each neuron, its output is calculated by the following formula:

[0117]

[0118] x j =f(S j ) (7)

[0119] Among them, f is the activation function, generally a sigmoid function or a linear function is selected.

[0120] The core of BP neural network (back propagation neural network) lies in its back propagation process, which is based on the gradient descent method and updates the weights and biases of the network by back propagating the error layer by layer, thereby gradually reducing the prediction error. In this process, the error signal starts from the output layer and propagates back to the input layer layer by layer, and the weights and biases in the network are adjusted according to these error signals.

[0121] First, we need to define a loss function (also called error function) E, which is used to quantify the difference between the network's predicted output and the actual output. The mean square error (MSE) is a common form of loss function. By calculating the square of the difference between the predicted value and the actual value and taking the average, we can get a numerical indicator that reflects the accuracy of the prediction. It can effectively measure the degree of deviation between the predicted value and the actual value. For a single sample, the mean square error loss function can be specifically defined as:

[0122]

[0123] Where M is the number of neurons in the output layer. aLj is the actual output of the jth neuron in the network output layer. yj is the corresponding target output (or actual output).

[0124] At this point, we need to determine an optimal set of weights and biases to minimize the value of the loss function E. To achieve this goal, we need to accurately calculate the partial derivatives of the loss function with respect to the weights and biases, that is, the gradients. For the weights and biases of the output layer, we can use a specific algorithm to calculate these gradient values:

[0125]

[0126] Where L represents the number of output layers;

[0127] is the weighted input of the jth neuron in the output layer;

[0128] is the output of the jth neuron in the output layer, where f is the activation function.

[0129] The derivation process of the weights and biases of the hidden layer and other layers is similar to that of the output layer, but the more complex chain rule needs to be considered. In a neural network, the output of each layer is the input of the next layer, so the output of the upper layer neurons will affect the input of the lower layer neurons. This dependency needs to be considered through the chain rule to ensure that the information flow of the entire network is accurately reflected when updating weights and biases. By continuously iteratively updating weights and biases, the predicted output of the network gradually approaches the actual output, thereby achieving accurate learning and prediction of data. Finally, the update formula for weights and biases can be expressed as:

[0130]

[0131] in, is the updated value of weight w; is the updated value of bias b; η is the learning rate, which is used to control the step size of parameter update and determines the amplitude and speed of parameter update. Too large a value may lead to unstable training, while too small a value may lead to too slow training. During the training process of the model, weights, biases and learning rates need to be adjusted to make the neural network better fit the training data.

[0132] 2.2 Prediction model variable selection

[0133] This model selects urban morphological variables such as residential population density, office population density, location advantage, road grade, parking facilities, commercial service facilities, daily life service facilities, sports and leisure service facilities, and traffic energy consumption values ​​as the input variables of the model. The output variable is the traffic energy consumption value of the road section calculated by the BP neural network.

[0134] In order to facilitate calculation, the present invention uniformly represents urban morphological variables such as population, parking, commerce, leisure facilities, and life service facilities with POI facilities, and links the number of POI facilities with the road network traffic energy consumption. At the same time, the location advantage is represented by the distance to the POI center point. Since different roads have different degrees of attraction to traffic flow, the road grade variable is also introduced into the model, and is divided into three levels: main roads, secondary roads, and branch roads for assignment. Then the input layer of the model contains three variables: the number of POI points, the location advantage value, and the road grade.

[0135] 2.3 Data Preprocessing

[0136] Data preprocessing is an indispensable part of improving model training efficiency. Since the original data may be mixed with various outliers, it cannot be directly input into the model for training. In order to ensure that the input value can fall within the response range of the S-type activation function and prompt the network to converge quickly, it is particularly important to preprocess the data. Therefore, outliers need to be processed before inputting variables to ensure that the data remains within the appropriate range.

[0137] 2.4 Selection of the number of hidden layers and neurons

[0138] A typical three-layer BP neural network structure covers an input layer, a hidden layer, and an output layer. Although the accuracy of the function can be improved to some extent by increasing the number of hidden layers, this will also increase the load on the system, prolong the training time, and increase the difficulty of model convergence. Under normal circumstances, a three-layer structure is sufficient to effectively map various functional relationships of general data. Therefore, this study selected a three-layer BP neural network to meet the model requirements, and the number of hidden layers selected was 1. As for the number of neurons in the hidden layer, there is no fixed method for determining it. In the present invention, the number of neurons in the hidden layer of the model is determined according to the degree of fit of the actual data. By continuously adjusting the number of neurons in the hidden layer and observing the changes in the fitting results of the model, the optimal model is selected. After many tests, it was found that when the number of neurons in the hidden layer was 8, the fitting result of the model was optimal. Therefore, the number of hidden layer neurons selected by the model is 8.

[0139] 2.5 Selecting the activation function

[0140] The purpose of introducing the activation function is to introduce nonlinear functions into the model. Since simple linear mapping is difficult to deal with complex data analysis, the activation function has become an effective tool to solve this problem. This model specifically uses the Sigmoid (logistic) activation function, which is also known as the S-shaped growth curve. It is a commonly used activation function, especially when applied to classifiers, its effect is more significant.

[0141] 2.6 Selecting the training function

[0142] The steepest descent BP algorithm (traingd) is selected to optimize the model. The algorithm first calculates the gradient vectors of each weight and threshold, which can indicate the direction in which the error surface drops most rapidly, and then uses the transfer function of the neural network to solve the gradient of the total error surface in multiple iterations. In each iteration, the steepest descent BP algorithm adjusts the weights and thresholds according to the calculated gradients to gradually reduce the total error. This process will continue until the error reaches the preset standard or meets other stopping conditions, which can effectively improve the prediction accuracy and overall performance of the model.

Claims

1. A method for predicting road traffic energy consumption in built-up areas of county towns, characterized in that: The following steps are involved: Step 1: Construct a road traffic energy consumption simulation model; Step 2: Construct a road traffic energy consumption prediction model to predict road traffic energy consumption.

2. The method for predicting road traffic energy consumption in built-up areas of county towns according to claim 1 is characterized in that: The step 1 specifically includes: Step 11: Divide the traffic energy consumption analysis area; divide the county built-up area into several sub-areas; when dividing, the overall block should be divided according to the road shape; Step 12: Determine the county center area; Step 13: Determine the urban morphology variable indicators; select the urban morphology indicators population density and POI facilities that have the greatest impact on urban road traffic volume as the urban morphology input variables of the traffic energy consumption simulation model; Step 14: Calculate location advantage; Step 15: Calculate traffic energy consumption; Step 16: Calculation of travel weight POI; Step 17: Analyze the distribution of traffic energy consumption in the area; Step 18: After calculating the energy consumption scores of each zone, the values ​​are distributed to the road network to obtain the traffic energy consumption value of each road section; Step 19: Use field measurements of road traffic volume to verify the accuracy of the model.

3. The method for predicting road traffic energy consumption in built-up areas of county towns according to claim 2 is characterized in that: The step 14 specifically includes: introducing a plot accessibility correction index to improve the traffic generation calculation of the plot unit; defining the plot unit accessibility to be determined by the planned road network density around the plot and the spatial distance between the plot and the central node within the planning range, as shown in the following formula: Where: K i is the accessibility of the analysis area i; P i P is the road network density around the analysis area i. According to the scale of the analysis area, the surrounding analysis area is appropriately selected to calculate its road network density; max is the road network density value of the analysis area with the maximum road network density within the research scope; L i is the spatial distance between the centroid of the analysis area i and the center point, which has been obtained through the standard ellipse difference analysis of the POI point in the above text; L min It is the minimum value of the distance between all analysis areas and the center point within the research scope.

4. The method for predicting road traffic energy consumption in built-up areas of county towns according to claim 2 is characterized in that: The step 15 specifically includes: designing a travel volume prediction model based on location advantage, population density, number of parking spaces, and POI facilities as follows: Among them, A i is the traffic volume of analysis area i; m represents the number of POI categories except residential and company categories; a q Represents the travel weight of the POI point of category q; S q represents the number of POI points of category q in the analysis area i; i represents the POI travel generation rate of the analysis area i; L represents the number of permanent residents in the study area; n represents the number of residential communities in the analysis area i; a r represents the travel incidence rate based on residential areas; L r represents the number of residents in the rth residential area in the analysis area; a j represents the travel incidence rate based on office companies; L j represents the number of employees in the jth company in the analysis area; K i is the locational advantage of analysis area i; Since the amount of traffic energy consumption generated is positively correlated with the travel volume, the traffic energy consumption attraction weight of the analysis area is expressed as follows: Among them, T i is the traffic energy consumption attraction weight of analysis area i, A i is the traffic volume of analysis area i, n is the number of analysis areas; Using refined oil sales data as the total value of the county road network traffic energy consumption, the traffic energy consumption generated by the sub-area is as follows: P i =P×T i (4) Among them, P i is the traffic energy consumption generated in the analysis area i, P is the total sales volume of refined oil, T i is the traffic energy consumption attraction weight of analysis area i.

5. The method for predicting road traffic energy consumption in built-up areas of county towns according to claim 2 is characterized in that: The step 16 specifically includes: the ratio of the POI facilities in each analysis area to the traffic volume in the built-up area is defined as the POI travel generation rate, which can be expressed as the following formula: Among them, O i represents the POI travel generation rate of the analysis area i; m represents the number of POI categories except residential and company categories; a q Represents the travel weight of the POI point of category q; S q represents the number of POI points of category q in analysis area i; n represents the number of analysis areas.

6. The method for predicting road traffic energy consumption in built-up areas of county towns according to claim 1 is characterized in that: The step 2 specifically includes: Step 21: Determine the input variables and output variables of the neural network; Step 22: Select the variables of the prediction model. The input layer includes three variables: the number of POI points, the location advantage value, and the road grade. Step 23: Before entering the variables, outliers need to be processed to ensure that the data remains within the appropriate range; Step 24: Select a three-layer BP neural network to meet the model requirements, and select 1 as the number of hidden layers; Step 25: Select activation function; Step 26: Select the steepest descent BP algorithm to optimize the model.

7. The method for predicting road traffic energy consumption in built-up areas of county towns according to claim 1, characterized in that: The step 21 specifically includes: the algorithm formula of the BP neural network involves two processes: forward propagation and back propagation.

8. The method for predicting road traffic energy consumption in built-up areas of county towns according to claim 7 is characterized in that: In the forward propagation process, data goes from the input layer through the hidden layer to the output layer; let the weight between node i and node i be w ij , the threshold of the node is b j , the output value of each node is x j , then for each neuron, its output is calculated by the following formula: x j =f(S j ) (6) Among them, f is the activation function.

9. The method for predicting road traffic energy consumption in built-up areas of county towns according to claim 7, characterized in that: The backpropagation process is based on the gradient descent method. It updates the weights and biases of the network by backpropagating the error layer by layer, thereby gradually reducing the prediction error. In this process, the error signal starts from the output layer and propagates back to the input layer layer by layer, and the weights and biases in the network are adjusted according to these error signals.

10. The method for predicting road traffic energy consumption in built-up areas of county towns according to claim 9, characterized in that: Define a loss function E to quantify the difference between the network prediction output and the actual output; mean square error MSE is a common form of loss function. By calculating the square of the difference between the predicted value and the actual value and finding the average, a numerical indicator reflecting the accuracy of the prediction is obtained, which can effectively measure the degree of deviation between the predicted value and the actual value; for a single sample, the mean square error loss function can be specifically defined as: Where M is the number of neurons in the output layer; is the actual output of the jth neuron in the network output layer; y j is the corresponding target output; Calculate the partial derivatives of the loss function with respect to the weights and biases, i.e. the gradients; for the weights and biases of the output layer, a specific algorithm is used to calculate these gradient values: Where L represents the number of output layers; is the weighted input of the jth neuron in the output layer; is the output of the jth neuron in the output layer, where f is the activation function; The update formula for weights and biases is expressed as: in, is the updated value of weight w; is the updated value of bias b; η is the learning rate.

Images