Urban vehicle order demand prediction method and device, equipment and storage medium

Abstract

The embodiment of the invention provides an urban vehicle order demand prediction method and device, equipment and a storage medium. The method comprises the following steps: dividing a city into a plurality of grid regions under a preset spatial resolution so as to obtain a city vehicle order demand OD matrix in a plurality of time steps; taking the urban vehicle order demand OD matrix in the previous T time steps as a sample, taking the urban vehicle order demand OD matrix in the next time step as a label, and constructing a data set; calculating the demand similarity between any two grid regions so as to construct a demand similarity matrix; according to the demand similarity matrix, constructing an irregular convolutional long-short term memory neural network; and training the constructed irregular convolutional long-short term memory neural network by using the data set to obtain an urban vehicle order demand prediction model, thereby accurately predicting the urban vehicle order demand based on the model, and improving the prediction precision.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular to a method, device, equipment and storage medium for predicting urban vehicle order demand. Background Art

[0002] With the improvement of computing power, machine learning and deep learning models are increasingly being used to predict urban vehicle order demand, such as support vector machines, random forests, Bayesian networks, Markov models, neural networks, and hybrid deep learning models. In particular, deep learning models have attracted widespread attention in the past decade, and many models have been used to predict vehicle order demand, such as convolutional neural networks, recurrent neural networks and their variants, encoder-decoders and attention mechanisms for sequential prediction, and graph convolutional networks for graph knowledge learning. However, most of these solutions ignore the demand correlation between regions and generally have poor prediction accuracy. Summary of the invention

[0003] In a first aspect, an embodiment of the present invention provides a method for training a city vehicle order demand prediction model, the method comprising:

[0004] The city is divided into multiple grid areas at a preset spatial resolution to obtain the OD matrix of urban vehicle order demand within multiple time steps;

[0005] The OD matrix of urban vehicle order demand in the previous T time steps is used as a sample, and the OD matrix of urban vehicle order demand in the next time step is used as a label to construct a data set;

[0006] Calculate the demand similarity between any two grid areas to construct a demand similarity matrix;

[0007] According to the required similarity matrix, an irregular convolutional long short-term memory neural network is constructed;

[0008] The dataset is used to train the constructed irregular convolutional long short-term memory neural network to obtain the urban vehicle order demand prediction model.

[0009] In some implementations of the first aspect, the OD matrix of city vehicle order demand in the previous T time steps is used as a sample, and the OD matrix of city vehicle order demand in the next time step is used as a label, before constructing the data set, the method further includes:

[0010] The obtained urban vehicle order demand OD matrix is ​​preprocessed and normalized.

[0011] In some implementations of the first aspect, calculating the demand similarity between any two grid areas includes:

[0012] For any two grid areas, the Euclidean distance between the centers of the two grid areas is calculated, and the demand similarity between the two grid areas is calculated based on this distance.

[0013] In some implementations of the first aspect, an irregular convolutional long short-term memory neural network is constructed according to the required similarity matrix, including:

[0014] For any grid area, the demand similarity corresponding to the grid area in the demand similarity matrix is ​​used as the weight parameter, and K grid areas are selected from multiple grid areas to construct their corresponding convolution kernels; according to the convolution kernels corresponding to each grid area, an irregular convolutional long short-term memory neural network is constructed.

[0015] In some implementations of the first aspect, the irregular convolutional long short-term memory neural network constructed is trained using a data set to obtain a city vehicle order demand prediction model, including:

[0016] Divide the data set into training set, validation set, and test set according to the preset ratio;

[0017] The training set is used to train the irregular convolutional long short-term memory neural network. During the training process, the validation set is used to verify the irregular convolutional long short-term memory neural network. After the training, the test set is used to test the irregular convolutional long short-term memory neural network. The irregular convolutional long short-term memory neural network that passes the test is used as a model for predicting urban vehicle order demand.

[0018] The irregular convolutional long short-term memory neural network constructed using the data set is trained to obtain a city vehicle order demand prediction model, including:

[0019] Divide the data set into training set, validation set, and test set according to the preset ratio;

[0020] The training set is used to train the irregular convolutional long short-term memory neural network. During the training process, the validation set is used to verify the irregular convolutional long short-term memory neural network. After the training, the test set is used to test the irregular convolutional long short-term memory neural network. The irregular convolutional long short-term memory neural network that passes the test is used as a model for predicting urban vehicle order demand.

[0021] In a second aspect, an embodiment of the present invention provides a method for predicting urban vehicle order demand, the method comprising:

[0022] Get the city vehicle order demand OD matrix within the first T time steps;

[0023] The obtained urban vehicle order demand OD matrix is ​​input into the urban vehicle order demand prediction model, and the urban vehicle order demand prediction model processes the input urban vehicle order demand OD matrix to predict the urban vehicle order demand OD matrix within the next time step;

[0024] Among them, the urban vehicle order demand prediction model is obtained based on the urban vehicle order demand prediction model training method as described above.

[0025] In a third aspect, an embodiment of the present invention provides a device for training a city vehicle order demand prediction model, the device comprising:

[0026] An acquisition module is used to divide the city into multiple grid areas at a preset spatial resolution, so as to obtain the OD matrix of the city's vehicle order demand within multiple time steps;

[0027] A construction module is used to construct a data set using the OD matrix of urban vehicle order demand in the previous T time steps as samples and the OD matrix of urban vehicle order demand in the next time step as labels;

[0028] The construction module is also used to calculate the demand similarity between any two grid areas to construct the demand similarity matrix;

[0029] The building module is also used to construct an irregular convolutional long short-term memory neural network based on the required similarity matrix;

[0030] The training module is used to train the constructed irregular convolutional long short-term memory neural network using the data set to obtain a city vehicle order demand prediction model.

[0031] In a fourth aspect, an embodiment of the present invention provides a device for predicting urban vehicle order demand, the device comprising:

[0032] The acquisition module is used to obtain the OD matrix of the city vehicle order demand within the previous T time steps;

[0033] A prediction module is used to input the acquired urban vehicle order demand OD matrix into the urban vehicle order demand prediction model, and the urban vehicle order demand prediction model processes the input urban vehicle order demand OD matrix to predict the urban vehicle order demand OD matrix within the next time step;

[0034] Among them, the urban vehicle order demand prediction model is obtained based on the urban vehicle order demand prediction model training method as described above.

[0035] In a fifth aspect, an embodiment of the present invention provides an electronic device, comprising: at least one processor; and a memory communicatively connected to the at least one processor; the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the method as described above.

[0036] In a sixth aspect, an embodiment of the present invention provides a non-transitory computer-readable storage medium storing computer instructions, where the computer instructions are used to enable a computer to execute the method described above.

[0037] In an embodiment of the present invention, the demand correlation between grid areas can be used to construct an irregular convolutional long short-term memory neural network. By training it, a city vehicle order demand prediction model with strong prediction ability can be obtained, and then based on the model, the city vehicle order demand can be accurately predicted to improve the prediction accuracy.

[0038] It should be understood that the contents described in the summary of the invention are not intended to limit the key or important features of the embodiments of the present invention, nor are they intended to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] The above and other features, advantages and aspects of the embodiments of the present invention will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. The accompanying drawings are provided for a better understanding of the present invention and do not constitute a limitation of the present invention. In the accompanying drawings, the same or similar reference numerals represent the same or similar elements, wherein:

[0040] Figure 1 A flow chart of a method for training a city vehicle order demand prediction model provided by an embodiment of the present invention;

[0041] Figure 2 A flow chart of a method for predicting urban vehicle order demand provided by an embodiment of the present invention;

[0042] Figure 3 A structural diagram of a city vehicle order demand forecasting model training device provided by an embodiment of the present invention;

[0043] Figure 4 A structural diagram of a device for predicting urban vehicle order demand provided by an embodiment of the present invention;

[0044] Figure 5 The figure is a structural diagram of an exemplary electronic device capable of implementing an embodiment of the present invention. DETAILED DESCRIPTION

[0045] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0046] In addition, the term "and / or" in the present invention is only a description of the association relationship of the associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in the present invention generally indicates that the associated objects before and after are in an "or" relationship.

[0047] In order to solve the technical problems arising from the background technology, the embodiments of the present invention provide a method, device, equipment and storage medium for predicting urban vehicle order demand. Specifically, the city is divided into multiple grid areas at a preset spatial resolution to obtain the urban vehicle order demand OD matrix within multiple time steps; the urban vehicle order demand OD matrix within the previous T time steps is used as a sample, and the urban vehicle order demand OD matrix within the next time step is used as a label to construct a data set; the demand similarity between any two grid areas is calculated to construct a demand similarity matrix; according to the demand similarity matrix, an irregular convolutional long short-term memory neural network is constructed; the constructed irregular convolutional long short-term memory neural network is trained using the data set to obtain an urban vehicle order demand prediction model, and then the urban vehicle order demand is predicted based on the model.

[0048] In this way, the demand correlation between grid areas can be used to construct an irregular convolutional long short-term memory neural network. By training it, a city vehicle order demand prediction model with strong prediction ability can be obtained, and then the city vehicle order demand can be accurately predicted based on the model to improve the prediction accuracy.

[0049] In conjunction with the accompanying drawings, a method, device, equipment and storage medium for predicting urban vehicle order demand provided by an embodiment of the present invention are described in detail through specific embodiments.

[0050] Figure 1 A flowchart of a method for training a city vehicle order demand prediction model provided by an embodiment of the present invention is shown in FIG. Figure 1 As shown, the urban vehicle order demand prediction model training method 100 may include the following steps:

[0051] S110, dividing the city into a plurality of grid areas at a preset spatial resolution, so as to obtain an OD matrix of the city's vehicle order demand within a plurality of time steps.

[0052] In some embodiments, the city may be divided into a plurality of regular grid areas at a spatial resolution of 1 km×1 km, and then the city vehicle order demand OD matrix within a plurality of time steps may be obtained based on the divided plurality of grid areas. The city vehicle order demand OD matrix may be defined as:

[0053] X od =(x ij ) n×n (1)

[0054] Where, X od represents the OD matrix of urban vehicle order demand; x ij Represents X od The element in the i-th row and j-th column in , that is, the number of orders that move from the i-th grid area to the j-th grid area within the time step; n represents the number of grid areas.

[0055] S120, constructing a data set using the OD matrix of city vehicle order demand in the previous T time steps as samples and the OD matrix of city vehicle order demand in the next time step as labels.

[0056] In some embodiments, the obtained city vehicle order demand OD matrix needs to be preprocessed and normalized first, and then the city vehicle order demand OD matrix within the previous T time steps is used as a sample and the city vehicle order demand OD matrix within the next time step is used as a label to construct a data set.

[0057] S130, calculating the demand similarity between any two grid areas, and constructing a demand similarity matrix.

[0058] In some embodiments, for any two grid areas, the Euclidean distance between the centers of the two grid areas can be calculated, and the demand similarity between the two grid areas can be calculated based on this distance, and then a demand similarity matrix can be constructed based on the demand similarity between the two grid areas. The elements in the demand similarity matrix can be defined as:

[0059]

[0060] In the formula, w ij represents the element in the i-th row and j-th column of the demand similarity matrix, that is, the demand similarity between grid area i and grid area j; d ij represents the Euclidean distance between the center of grid area i and the center of grid area j, σ 2 , ε is a hyperparameter for adjusting the demand similarity matrix.

[0061] S140, constructing an irregular convolutional long short-term memory neural network according to the required similarity matrix.

[0062] In some embodiments, for any grid area, the demand similarity corresponding to the grid area in the demand similarity matrix can be used as a weight parameter, K grid areas can be randomly selected from multiple grid areas to construct their corresponding convolution kernels, and then an irregular convolution long short-term memory neural network is constructed according to the convolution kernels corresponding to each grid area. The constructed irregular convolution long short-term memory neural network can be expressed using the following formula:

[0063] i t =σ(W xi *X t +W hi *H t―1 +b i ) (3)

[0064] f t =σ(W xf *X t +W hf *H t―1 +b f ) (4)

[0065]

[0066] o t =σ(W xo *X t +W ho *H t―1 +b o ) (6)

[0067]

[0068] Among them, * represents the irregular convolution operation. In the * process, the convolution kernel of each area is pre-built based on the above operation.

[0069] S150, using the data set to train the constructed irregular convolutional long short-term memory neural network to obtain a city vehicle order demand prediction model.

[0070] In some embodiments, the data set can be divided into a training set, a validation set, and a test set according to a preset ratio (for example, 3:1:1), and then the training set is used to train the irregular convolutional long short-term memory neural network. The validation set is used to verify the irregular convolutional long short-term memory neural network during the training process. After the training, the test set is used to test the irregular convolutional long short-term memory neural network, and finally the irregular convolutional long short-term memory neural network that passes the test is used as a city vehicle order demand prediction model.

[0071] Based on the urban vehicle order demand prediction model training method 100 provided in the embodiment of the present invention, the embodiment of the present invention also provides an urban vehicle order demand prediction method, such as Figure 2 As shown, the city vehicle order demand forecasting method 200 may include the following steps:

[0072] S210, obtaining the city vehicle order demand OD matrix within the first T time steps.

[0073] S220, input the obtained urban vehicle order demand OD matrix into the urban vehicle order demand prediction model, and the urban vehicle order demand prediction model processes the input urban vehicle order demand OD matrix to predict the urban vehicle order demand OD matrix within the next time step.

[0074] Among them, the urban vehicle order demand prediction model is obtained based on the urban vehicle order demand prediction model training method as described above.

[0075] The following is a comparison of the performance of LSTM, BI-LSTM, ConvLSTM and the irregular convolutional long short-term memory neural network in the urban vehicle order demand prediction task to intuitively reflect the superiority of the present invention, as shown below:

[0076] LSTM is selected as a weak learning model for predicting urban vehicle order demand. Among them, LSTM, as a recurrent neural network algorithm, can effectively solve the problem of time series modeling. LSTM effectively solves the gradient vanishing problem of the original recurrent neural network by adding different state gates, namely input gate, forget gate and output gate. The LSTM here includes 1 LSTM layer and 1 fully connected layer. Its input dimension needs to be determined by the sample dimension of the urban vehicle order demand OD matrix and is expressed as T×N×N. The number of hidden layer unit nodes is set to N×N, and the size of the output layer is N×N, that is, the size of the urban vehicle order demand OD matrix.

[0077] BI-LSTM is selected as a weak learning model for predicting urban vehicle order demand. As an improved recurrent neural network algorithm, BI-LSTM can capture the bidirectional dependency in the sequence by combining the outputs of the two LSTMs in the forward and backward directions. The BI-LSTM here includes 1 BI-LSTM layer and 1 fully connected layer. Its input dimension needs to be determined by the sample dimension of the urban vehicle order demand OD matrix and is expressed as T×N×N. The number of hidden layer unit nodes is set to 2×N×N, and the size of the output layer is N×N.

[0078] ConvLSTM is selected as a weak learning model for predicting urban vehicle order demand. As an improved neural network algorithm, ConvLSTM not only inherits the advantages of traditional LSTM, but also makes it very suitable for spatiotemporal data due to its internal convolutional structure. The ConvLSTM here includes 1 ConvLSTM layer and 1 fully connected layer. Its input dimension needs to be determined by the sample dimension of the urban vehicle order demand OD matrix and is expressed as T×N×N. The number of hidden layer unit nodes is set to N×N, and the size of the output layer is N×N.

[0079] Here we choose Manhattan, New York, which is a common object in the study of road network complexity. The Manhattan road network has a regular shape, with streets crisscrossed horizontally and vertically, and its road network dataset is easy to obtain through the New York City taxi data open source website.

[0080] Based on the above settings, the prediction results of the urban vehicle order demand OD matrix in the Manhattan road network by LSTM, BI-LSTM, ConvLSTM and the irregular convolutional long short-term memory neural network in the present invention are compared here, as shown in Table 1:

[0081] Table 1

[0082]

[0083] In summary, according to the embodiments of the present invention, at least the following technical effects are achieved:

[0084] (1) The present invention can effectively improve the generalization ability of the prediction model by means of a neural network mechanism. Compared with existing neural network learning or deep learning models, the model application in the present invention has higher generalization.

[0085] (2) The present invention utilizes the demand correlation between regions to ensure that the neural network can ensure the classification and regression accuracy while improving the nonlinear learning ability of data samples and the adaptability to the spatiotemporal variation characteristics. To a certain extent, the nonlinear learning ability of the model is improved to meet the training needs of the regression task. Therefore, it expands the sample range to a certain extent, thereby increasing the robustness of the model.

[0086] (3) In practical applications, the present invention does not need to perform other feature processing on the urban vehicle order demand OD matrix samples in advance. The method can directly achieve end-to-end training output and has lower requirements on the environment.

[0087] It should be noted that, for the above-mentioned method embodiments, for the sake of simplicity, they are all described as a series of action combinations, but those skilled in the art should know that the present invention is not limited by the described order of actions, because according to the present invention, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all optional embodiments, and the actions and modules involved are not necessarily required by the present invention.

[0088] The above is an introduction to a method embodiment. The following is a further explanation of the solution of the present invention through an apparatus embodiment.

[0089] Figure 3 A structural diagram of a city vehicle order demand forecasting model training device provided by an embodiment of the present invention, such as Figure 3 As shown, the urban vehicle order demand prediction model training device 300 may include:

[0090] The acquisition module 310 is used to divide the city into multiple grid areas at a preset spatial resolution, so as to obtain the city vehicle order demand OD matrix within multiple time steps.

[0091] The construction module 320 is used to construct a data set using the OD matrix of city vehicle order demand in the previous T time steps as samples and the OD matrix of city vehicle order demand in the next time step as labels.

[0092] The construction module 320 is further used to calculate the demand similarity between any two grid areas, so as to construct a demand similarity matrix.

[0093] The construction module 320 is also used to construct an irregular convolutional long short-term memory neural network according to the required similarity matrix.

[0094] The training module 330 is used to train the constructed irregular convolutional long short-term memory neural network using the data set to obtain a city vehicle order demand prediction model.

[0095] Understandably, Figure 3 Each module / unit in the urban vehicle order demand prediction model training device 300 has the following functions: Figure 1 The functions of each step in the urban vehicle order demand prediction model training method 100 shown in the figure can achieve its corresponding technical effects, and for the sake of brevity, they will not be repeated here.

[0096] Figure 4 A structural diagram of a city vehicle order demand forecasting device provided by an embodiment of the present disclosure, such as Figure 4 As shown, the urban vehicle order demand forecasting device 400 may include:

[0097] The acquisition module 410 is used to obtain the city vehicle order demand OD matrix within the first T time steps.

[0098] The prediction module 420 is used to input the acquired urban vehicle order demand OD matrix into the urban vehicle order demand prediction model, and the urban vehicle order demand prediction model processes the input urban vehicle order demand OD matrix to predict the urban vehicle order demand OD matrix within the next time step.

[0099] Among them, the urban vehicle order demand prediction model is obtained based on the urban vehicle order demand prediction model training method as described above.

[0100] Understandably, Figure 4 Each module / unit in the urban vehicle order demand forecasting device 400 shown in FIG. Figure 2 The functions of each step in the urban vehicle order demand forecasting method 200 shown in the figure can achieve its corresponding technical effects, and for the sake of brevity, they will not be repeated here.

[0101] Figure 5 5 is a block diagram of an exemplary electronic device capable of implementing an embodiment of the present invention. Electronic device 500 is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. Electronic device 500 may also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown in the present invention, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or required in the present invention.

[0102] like Figure 5 As shown, the electronic device 500 may include a computing unit 501, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 502 or a computer program loaded from a storage unit 508 into a random access memory (RAM) 503. In the RAM 503, various programs and data required for the operation of the electronic device 500 can also be stored. The computing unit 501, the ROM 502, and the RAM 503 are connected to each other via a bus 504. An input / output (I / O) interface 505 is also connected to the bus 504.

[0103] Multiple components in the electronic device 500 are connected to the I / O interface 505, including: an input unit 506, such as a keyboard, a mouse, etc.; an output unit 507, such as various types of displays, speakers, etc.; a storage unit 508, such as a disk, an optical disk, etc.; and a communication unit 509, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 509 allows the electronic device 500 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0104] The computing unit 501 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the computing unit 501 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, digital signal processors (DSPs), and any appropriate processors, controllers, microcontrollers, etc. The computing unit 501 performs the various methods and processes described above, such as method 100 or method 200. For example, in some embodiments, method 100 or method 200 may be implemented as a computer program product, including a computer program, which is tangibly contained in a computer-readable medium, such as a storage unit 508. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 500 via ROM 502 and / or communication unit 509. When the computer program is loaded into RAM 503 and executed by the computing unit 501, one or more steps of the method 100 or method 200 described above may be performed. Alternatively, in other embodiments, the computing unit 501 may be configured to execute the method 100 or the method 200 in any other appropriate manner (eg, by means of firmware).

[0105] The various embodiments described above in the present invention can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include: being implemented in one or more computer programs, which may be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general programmable processor, which may receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0106] The program code for implementing the method of the present invention can be written in any combination of one or more programming languages. These program codes can be provided to a processor or controller of a general-purpose computer, a special-purpose computer or other programmable data processing device, so that the program code, when executed by the processor or controller, enables the functions / operations specified in the flow chart and / or block diagram to be implemented. The program code can be executed entirely on the machine, partially on the machine, partially on the machine as a stand-alone software package and partially on a remote machine, or entirely on a remote machine or server.

[0107] In the context of the present invention, a computer-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A computer-readable medium may be a computer-readable signal medium or a computer-readable storage medium. A computer-readable medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or equipment, or any suitable combination of the foregoing. A more specific example of a computer-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0108] It should be noted that the present invention also provides a non-transitory computer-readable storage medium storing computer instructions, wherein the computer instructions are used to enable a computer to execute method 100 or method 200 and achieve the corresponding technical effects achieved by executing the methods in the embodiments of the present invention. For the sake of concise description, they will not be repeated here.

[0109] In addition, the present invention also provides a computer program product, which includes a computer program. When the computer program is executed by a processor, the method 100 or the method 200 is implemented.

[0110] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps described in the present invention can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solution disclosed in the present invention can be achieved, and the present invention is not limited here.

[0111] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for training a city vehicle order demand prediction model, characterized in that: The method comprises: The city is divided into multiple grid areas at a preset spatial resolution to obtain the OD matrix of urban vehicle order demand within multiple time steps; The OD matrix of urban vehicle order demand in the previous T time steps is used as a sample, and the OD matrix of urban vehicle order demand in the next time step is used as a label to construct a data set; Calculate the demand similarity between any two grid areas to construct a demand similarity matrix; According to the demand similarity matrix, an irregular convolutional long short-term memory neural network is constructed; The dataset is used to train the constructed irregular convolutional long short-term memory neural network to obtain a city vehicle order demand prediction model.

2. The method according to claim 1, characterized in that: The method further includes, before constructing the data set, using the city vehicle order demand OD matrix within the previous T time steps as a sample and the city vehicle order demand OD matrix within the next time step as a label: The obtained urban vehicle order demand OD matrix is ​​preprocessed and normalized.

3. The method according to claim 1, characterized in that The calculating of the demand similarity between any two grid areas includes: For any two grid areas, the Euclidean distance between the centers of the two grid areas is calculated, and the demand similarity between the two grid areas is calculated based on this distance.

4. The method according to claim 1, characterized in that: The step of constructing an irregular convolutional long short-term memory neural network according to the demand similarity matrix includes: For any grid area, the demand similarity corresponding to the grid area in the demand similarity matrix is ​​used as a weight parameter, and K grid areas are selected from multiple grid areas to construct their corresponding convolution kernels; according to the convolution kernels corresponding to each grid area, an irregular convolution long short-term memory neural network is constructed.

5. The method according to claim 1, characterized in that: The method uses the data set to train the constructed irregular convolutional long short-term memory neural network to obtain a city vehicle order demand prediction model, including: Dividing the data set into a training set, a validation set, and a test set according to a preset ratio; The irregular convolutional long short-term memory neural network is trained using the training set, the irregular convolutional long short-term memory neural network is verified using the verification set during the training process, and the irregular convolutional long short-term memory neural network is tested using the test set after the training is completed, and the irregular convolutional long short-term memory neural network that passes the test is used as a city vehicle order demand prediction model.

6. A method for predicting urban vehicle order demand, characterized in that: The method comprises: Get the city vehicle order demand OD matrix within the first T time steps; The obtained city vehicle order demand OD matrix is ​​input into the city vehicle order demand prediction model, and the city vehicle order demand prediction model processes the input city vehicle order demand OD matrix to predict the city vehicle order demand OD matrix within the next time step; Wherein, the urban vehicle order demand forecasting model is obtained based on the urban vehicle order demand forecasting model training method described in any one of claims 1-5.

7. A training device for a city vehicle order demand prediction model, characterized in that: The device comprises: An acquisition module is used to divide the city into multiple grid areas at a preset spatial resolution, so as to obtain the OD matrix of the city's vehicle order demand within multiple time steps; A construction module is used to construct a data set using the OD matrix of urban vehicle order demand in the previous T time steps as samples and the OD matrix of urban vehicle order demand in the next time step as labels; The construction module is also used to calculate the demand similarity between any two grid areas, so as to construct a demand similarity matrix; The construction module is further used to construct an irregular convolutional long short-term memory neural network according to the demand similarity matrix; The training module is used to train the constructed irregular convolutional long short-term memory neural network using the data set to obtain a city vehicle order demand prediction model.

8. A device for predicting urban vehicle order demand, characterized in that: The device comprises: The acquisition module is used to obtain the OD matrix of the city vehicle order demand within the previous T time steps; A prediction module, used for inputting the acquired city vehicle order demand OD matrix into a city vehicle order demand prediction model, wherein the city vehicle order demand prediction model processes the input city vehicle order demand OD matrix and predicts the city vehicle order demand OD matrix within the next time step; Wherein, the urban vehicle order demand forecasting model is obtained based on the urban vehicle order demand forecasting model training method described in any one of claims 1-5.

9. An electronic device, characterized in that: The electronic device comprises: at least one processor; and a memory communicatively connected to the at least one processor; wherein, The memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can perform the method according to any one of claims 1 to 6.

10. A non-transitory computer-readable storage medium storing computer instructions, characterized in that: The computer instructions are used to enable a computer to execute the method according to any one of claims 1 to 6.

Images