Electric vehicle charging scheduling method based on time cost and micro-grid load balance

Abstract

The invention discloses an electric vehicle charging scheduling method based on time cost and micro-grid load balance. The method comprises the following steps: step 1, user time cost calculation; step 2, updating the load of the MG system in the time period from charging start to charging end of the electric vehicle, and then obtaining the average load of the MG system according to the updated load of the MG system and the charging duration; and step 3, obtaining a total index of the # imgabs0 # vehicle on the # imgabs1 # charging station through normalization and weighting, and giving an optimal charging scheduling strategy through an MTC-SLBMS algorithm by taking the total index as an input. The method effectively reduces the charging time cost of the EV user, and optimizes the charging experience of the user. And meanwhile, the space load balance between MMGs is realized, and the operation safety of the charging station is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of electric vehicle charging, and in particular to an electric vehicle charging scheduling method based on time cost and microgrid load balancing. Background Art

[0002] In recent years, the global electric vehicle (EV) industry has shown a rapid development trend, and EV has become an important choice to replace traditional fuel vehicles. The rapid development of EV has brought significant impacts on the power system, especially in the load management of multi-microgrids (MMGs). Electric vehicles will generate higher power demand when charging in a centralized manner, which causes MMGs to bear greater load pressure in a short period of time. Especially in some urban areas with a high EV penetration rate, the problem of uneven load faced by MMGs is becoming increasingly prominent. Uneven load may cause local MG overload and affect power supply reliability. Therefore, effective solutions are needed to manage and balance the load of MMGs. If the charging behavior of EV users can be guided by effective strategies, EVs can become a tool for balancing the load between MMGs, which will greatly reduce the impact of centralized EV charging on MMGs. By formulating a new charging scheduling algorithm that meets the needs of EV users and balances the load differences between MMGs, the efficiency and reliability of the EV charging process can be ensured.

[0003] Specifically, the prior art [1] Y. Xiang, J. Yang, X. Li, C. Gu, S. Zhang, Routing optimization of electric vehicles for charging with event-driven pricing strategy, IEEE Transactions on Automation Science and Engineering 19 (1)(2021) 7– 20. discloses the optimal charging route planning for a single vehicle based on a road network model. In addition, [2] X. Li, Y. Xiang, L. Lyu, C. Ji, Q. Zhang, F. Teng, Y. Liu, Price incentive-based charging navigation strategy for electric vehicles, IEEE Transactions on Industry Applications 56 (5) (2020) 5762–5774 proposed a large-scale vehicle scheduling algorithm that minimizes the user charging time cost based on the literature [1]. However, both of them did not take into account the impact of large-scale vehicles on MMG load.

[0004] In addition, [3] X. Chen, H. Wang, F. Wu, Y. Wu, MC Gonz´alez, J. Zhang, Multi-microgrid load balancing through EV charging networks, IEEE Internet of Things Journal 9 (7) (2021) 5019–5026. A spatial load balancing algorithm for large-scale vehicle charging scheduling is proposed, but it does not take into account the user's time cost, resulting in high queuing time costs for users at local charging stations. Therefore, a new scheduling algorithm is proposed from both the user's perspective and the MMG's perspective to jointly optimize the user's time cost and the MMG spatial load balancing degree. Summary of the invention

[0005] In view of the shortcomings of the prior art, the present invention proposes an electric vehicle charging scheduling method based on time cost and microgrid load balancing. By integrating real-time traffic information, charging station information and MMG load information, the present invention accurately formulates charging navigation strategies for users and reasonably regulates the number of EVs accepted by each charging station. It effectively reduces the charging time cost of EV users and optimizes the user charging experience. At the same time, it realizes spatial load balancing between MMGs and improves the operating safety of charging stations.

[0006] In order to solve the above technical problems, the technical solution of the present invention is:

[0007] A method for scheduling electric vehicle charging based on time cost and microgrid load balancing includes the following steps:

[0008] Step 1: Calculate the user time cost, which is the sum of the travel time cost of the electric vehicle to each charging station, the waiting time cost of the corresponding charging station, and the charging time cost;

[0009] Step 2: Determine the minimum and maximum loads of the MMG system at a certain moment, and give priority to charging electric vehicles at charging stations with lower loads. The electric car is heading to The electric vehicle is charged at a charging station, and the load of the MG system during the period from the start of charging to the end of charging is updated. Then, the average load of the MG system is obtained according to the updated load of the MG system and the charging time.

[0010] Step 3: Combine the user time cost obtained in step 1 and the average load of the MG system obtained in step 2, and obtain the first The car is in The total index of the charging stations is used as input to give the optimal charging scheduling strategy through the MTC-SLBMS algorithm.

[0011] Preferably, in step 1, the method for estimating the travel time cost of the electric vehicle to each charging station is:

[0012] First, the traffic network is modeled. The traffic network model is expressed as ,in is the set of all road nodes, is the set of all connected segments, is the set of road weights,

[0013] When Electric vehicles in When a charging request is issued, the initial power state is , the electric vehicle dispatch center will record the The location of the electric car, Electric vehicles on the road Driving speed It can be expressed as:

[0014]

[0015]

[0016] in, Road Section The free flow speed, Indicates road segment The traffic capacity, Road Section exist Traffic flow at any time, and The ratio is the road segment Road saturation at the moment, , , is the adaptive coefficient for different road levels, and the speed-flow model is used to calculate the driving speed of electric vehicles on different road sections. , the driving time of electric vehicles on each road section is calculated by the following formula:

[0017]

[0018] in Represents a slave node To Node distance, so that a time-weighted matrix can be constructed ,

[0019]

[0020] in Represents a slave node To Node Estimated travel time.

[0021] Preferably, in step 1, the method for estimating the travel time cost of the electric vehicle to each charging station further comprises:

[0022] Using the Floyd algorithm, substitute the time weighted matrix , and finally generate the node sequence contained in the shortest path:

[0023]

[0024] in, Each node in represents a specific road node on the driving path of the electric vehicle. is the starting node, i.e. the current location of the electric vehicle, and is the terminal node, i.e. the location of the charging station;

[0025] Thus, the distance to the charging station in each MMG system is calculated. The distance between an electric vehicle and each charging station can be represented by a vector R, which is defined as follows:

[0026]

[0027] In the formula No. Electric cars to The distance between charging stations, N represents the number of electric vehicles in the MMG system, represents the number of charging stations in the MMG system, so Expressed as the sum of the lengths of the segments in the planned path:

[0028]

[0029] The speed-energy consumption model used reflects the relationship between energy consumption and driving speed:

[0030]

[0031] In the formula , , They are the energy consumption per unit mileage of urban expressways, main roads, and secondary roads;

[0032] Then find the Electric cars to Energy consumption per charging station :

[0033]

[0034] Determine the reach of the electric vehicle based on the driving energy consumption, determine whether the remaining SoC of the electric vehicle can support reaching the target charging station, and generate a candidate set ,Finally, determine the starting point of each electric vehicle user, using the updated time weighted matrix To traverse the shortest time path to each charging station, in the candidate set Recommend suitable charging stations for electric vehicles.

[0035] Preferably, in the traffic network model, the length of the road section is used as the weight of the traffic network, thereby obtaining a set of road weights: , the set of road weights Elements in It is expressed as:

[0036]

[0037] in, Indicates that from Click to The distance between points, assuming that the roads are bidirectionally connected, then .

[0038] Preferably, the candidate set Set constraints, candidate set The constraints are expressed as: .

[0039] Preferably, in step 1, the queuing time cost is calculated as follows:

[0040] When Electric vehicles at the time Submit a charging application and send the current electric vehicle location to the dispatch center. The traffic network model is used to calculate the time it takes to reach the Time on the road at each charging station , Electric vehicle arrival time It can be calculated as:

[0041]

[0042] The arrival time of electric vehicles at the charging station and the start time of charging meet the following requirements:

[0043]

[0044]

[0045] In the formula, Represents electric vehicle slave node Arrival at the charging station The waiting time, Indicates Charging station The estimated charging end time of each charging pile, For the Number of charging piles per charging station, Indicates The earliest charging end time of all charging piles in a charging station is calculated, and the waiting time is determined by calculating the difference between the earliest charging end time and the arrival time of the electric vehicle.

[0046] Preferably, in step 1, the charging time cost calculation method is:

[0047] Known Electric vehicles in The initial power state at time , then the electric car reaches the Charging Station Remaining SoC satisfy:

[0048]

[0049] in for Electric cars go to The energy consumption of each charging station is calculated, and the remaining SoC of the electric vehicle after arriving at the station is calculated. Substitute into the charging curve to calculate the charging time .

[0050] Preferably, the charging time and The relationship satisfies the following conditions:

[0051]

[0052]

[0053]

[0054] Traverse and find the estimated waiting time for each power station ,in j, then add the time cost of the electric vehicle on the way to each charging station, the corresponding waiting time cost and the charging time cost, and record it as the total time cost :

[0055] .

[0056] Preferably, the load updating method of the MG system is:

[0057]

[0058] in, For the MG system where charging stations are located The load value at the moment, for The initial load at time, Indicates The moment an electric car starts to connect to charging.

[0059] Preferably, in step 2, the average load is introduced to describe the load change of the MG system from the time when the electric vehicle is connected to the grid and starts charging to the time when the charging ends. Electric cars arrive When there are 1 charging station, the average load expression of the MG system is as follows:

[0060] .

[0061] Preferably, in step 3, after the user time cost and the average load of the MG system are normalized, the weight is calculated by the entropy weight method to obtain the first The car is in The total index of the charging stations, so as to select the optimal charging station according to the total index :

[0062]

[0063]

[0064]

[0065] in, and are the weight coefficients calculated by the entropy weight method.

[0066] The present invention has the following characteristics and beneficial effects:

[0067] By adopting the above technical solution, the present invention establishes an EV charging network model that couples MMG with a traffic network. From the dual perspectives of users and MMGs, the user's charging time cost and the load fluctuation of MMGs during the charging process are comprehensively considered. This embodiment proposes a minimum time cost-load balancing strategy (MTC-SLBMS). This strategy aims to optimize the user experience and MMG stability at the same time, and achieve a win-win situation for users and power grids through effective scheduling. In the simulation study, the MTC-SLBMS algorithm is compared with the traditional shortest path strategy (SDMS), the minimum time cost strategy (MTMS), the load balancing strategy (LBMS) and the improved load balancing strategy (ILBMS) in this embodiment. The results show that MTC-SLBMS can not only significantly reduce the user's charging time cost, but also effectively balance the load fluctuation of MMGs. This strategy not only retains the respective advantages of MTMS and ILBMS, but also overcomes their shortcomings and provides a more comprehensive solution. Specifically, MTC-SLBMS saves user time costs by dynamically adjusting the distribution of charging demand, while avoiding the problem of overload of the power grid during peak hours. Such a design not only improves the charging efficiency of EV users, but also enhances the stability and reliability of MMGs. The present invention provides a more comprehensive and efficient solution for large-scale electric vehicle charging scheduling, optimizing the overall system performance. BRIEF DESCRIPTION OF THE DRAWINGS

[0068] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0069] Figure 1 This is a flow chart of an electric vehicle charging scheduling method based on time cost and microgrid load balancing according to an embodiment of the present invention.

[0070] Figure 2 A schematic diagram of feasibility verification of the allocation strategy of an embodiment of the present invention.

[0071] Figure 3 This is a comparative diagram of how the user's time cost changes with the increase in the number of vehicles under different strategies.

[0072] Figure 4 Schematic diagram of the comparison of the average valley-to-peak ratio under different strategies as the number of vehicles increases. DETAILED DESCRIPTION

[0073] It should be noted that, in the absence of conflict, the embodiments of the present invention and the features in the embodiments may be combined with each other.

[0074] In the description of the present invention, it should be understood that the terms "center", "longitudinal", "lateral", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside" and the like indicate positions or positional relationships based on the positions or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, the terms "first", "second", and the like are only used for descriptive purposes, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, features defined as "first", "second", and the like may explicitly or implicitly include one or more of the features. In the description of the present invention, unless otherwise specified, "multiple" means two or more.

[0075] In the description of the present invention, it should be noted that, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood by specific circumstances.

[0076] The present invention provides an electric vehicle charging scheduling method based on time cost and microgrid load balancing. By establishing an EV charging network model that couples MMGs with traffic networks, the charging time cost of users and the load fluctuation of MMGs are comprehensively considered from the dual perspectives of users and MMGs, and a new scheduling algorithm MTC-SLBMS is proposed, which considers the impact of realistic factors such as traffic flow, queue waiting time and charging time on large-scale EV charging scheduling. At the same time, the spatial load balancing degree between MMGs is also considered, and the two indicators of user time cost and MMG power are integrated to achieve the purpose of joint optimization.

[0077] Specifically, it is achieved through the following methods:

[0078] Step 1: User time cost

[0079] (1) Road travel time cost

[0080] First, the traffic network is modeled. The traffic network model is expressed as ,in is the set of all road nodes, is the set of all connected segments, In this embodiment, the length of the road section is used as the weight of the traffic network in the constructed traffic network model, so that the generated road weight matrix set Elements in It is expressed as:

[0081]

[0082] in, Indicates that from Click to The distance between points, assuming that the roads are bidirectionally connected, then When the Electric vehicles in When a charging request is issued, the initial power state is , the electric vehicle dispatch center will record the The location of the electric car.

[0083] Understandably, in urban traffic networks, the speed of electric vehicles is mainly affected by road capacity and traffic flow. Electric vehicles on the road Driving speed It can be expressed as:

[0084]

[0085]

[0086] in, Road Section The free flow speed, Indicates road segment The traffic capacity, Road Section exist Traffic flow at any time, and The ratio is the road segment Road saturation at the moment, , , is the adaptive coefficient for different road levels.

[0087] It should be noted that in the speed-flow model, roads are divided into urban expressways, main roads and secondary roads. In this embodiment, the speed-flow model is used to calculate the driving speed of electric vehicles on different road sections. , the driving time of electric vehicles on each road section is calculated by the following formula:

[0088]

[0089] in Represents a slave node To Node distance, so that a time-weighted matrix can be constructed :

[0090]

[0091] in Represents a slave node To Node Estimated travel time.

[0092] Then use the Floyd algorithm and substitute the time weighted matrix , to find the shortest travel time path between any two points, and finally generate the node sequence contained in the shortest path:

[0093]

[0094] Among them, the sequence Each node in represents a specific road node on the driving path of the electric vehicle. The path passes through these nodes in sequence from the starting point to the end point. is the starting node, i.e. the current location of the electric vehicle, and is the terminal node, i.e. the location of the charging station.

[0095] Thus, the distance to the charging station in each MMG system is calculated. The distance between an electric vehicle and each charging station can be represented by a vector R, which is defined as follows:

[0096]

[0097] In the formula No. Electric cars to The distance between charging stations, N represents the number of electric vehicles in the MMG system, represents the number of charging stations in the MMG system, so Expressed as the sum of the lengths of the segments in the planned path:

[0098]

[0099] It is understandable that in an urban traffic network, the energy consumption per unit mileage of electric vehicles under different traffic conditions varies greatly. Therefore, in this embodiment, the speed-energy consumption model cited reflects the relationship between energy consumption and driving speed:

[0100]

[0101] In the formula , , They are the energy consumption per unit mileage of urban expressways, main roads, and secondary roads;

[0102] Then find the Electric cars to Energy consumption per charging station :

[0103]

[0104] Determine the reach of the electric vehicle based on the driving energy consumption, determine whether the remaining SoC of the electric vehicle can support reaching the target charging station, and generate a candidate set , where the candidate set The constraints can be expressed as:

[0105]

[0106] Finally, determine the starting point of each electric vehicle user, using the updated time-weighted matrix To traverse the shortest time path to each charging station, in the candidate set Recommend suitable charging stations for electric vehicles.

[0107] This is understandable because each vehicle generates the shortest path for its target charging station after using the algorithm to calculate the most suitable target charging station. In the case of large-scale vehicles, many vehicles are already traveling to their respective target charging stations before the vehicle being assigned at the current moment, so the traffic flow on the road will change, thus affecting the vehicle's travel time. Therefore, each vehicle will use the time-weighted matrix updated at the current moment when calculating.

[0108] (2) Queuing time cost

[0109] Understandably, when allocating a suitable charging station to a user, the waiting time in the queue at the power station and the charging time are also considered from the user's perspective. Taking into account the uncertainty of the time when EVs arrive and leave the charging station, the estimated waiting time after the vehicle arrives at the station is calculated by generating an estimated queue.

[0110] When Electric vehicles at the time Submit a charging application and send the current electric vehicle location to the dispatch center. The traffic network model is used to calculate the time it takes to reach the Time on the road at each charging station , Electric vehicle arrival time It can be calculated as:

[0111]

[0112] The arrival time of electric vehicles at the charging station and the start time of charging meet the following requirements:

[0113]

[0114]

[0115] In the formula, Represents electric vehicle slave node Arrival at the charging station The waiting time, Indicates Charging station The estimated charging end time of each charging pile, For the Number of charging piles per charging station, Indicates The earliest charging end time of all charging piles in a charging station is calculated, and the waiting time is determined by calculating the difference between the earliest charging end time and the arrival time of the electric vehicle.

[0116] (3) Charging time cost

[0117] Known Electric vehicles in The initial power state at time , then the electric car reaches the The remaining SoCs of charging stations meet the following requirements:

[0118]

[0119] in for Electric cars go to The energy consumed by each charging station is calculated, and the remaining SoC of the electric vehicle after arriving at the station is substituted into the charging curve to calculate the charging time. .

[0120] Charging time and The relationship satisfies the following conditions:

[0121]

[0122]

[0123]

[0124] Traverse and find the estimated waiting time for each power station ,in j, and then add the time cost of the electric vehicle on the way to each charging station, the corresponding waiting time cost and the charging time cost, and record it as the total time cost:

[0125] .

[0126] Step 2: Space load balancing

[0127] MMG system The minimum load and maximum load existing at any time are expressed as and Understandably, vehicles are given priority to go to charging stations with lower loads for charging, so as to achieve the purpose of "peak shaving and valley filling", stabilize the MMG power, and ensure the safety of MMG operation.

[0128] Assume Electric cars go to The load update of MG during the period from the start of charging to the end of charging can be expressed as:

[0129]

[0130] in, For the MG system where charging stations are located The load value at the moment, for Initial load at time, Indicates the time when the first electric vehicle starts to be connected for charging. Indicates EV in the first Charging time at each charging station.

[0131] The MG system refers to a single MG where a vehicle is charged. In this embodiment, a charging station is regarded as a MG system.

[0132] It should be noted that when formulating charging distribution strategies, future load changes must be considered comprehensively, rather than based solely on current or past instantaneous data. To describe the MG load change from the time the vehicle is connected to the grid and starts charging to the time the charging ends. Electric cars arrive at The average load of the MG system when there are charging stations Defined as:

[0133]

[0134] Average load The smaller the During EV charging The lower the average load of each MG, the more MGs are charged. This fully estimates the change in MG load from the time EV is connected to the grid to the time EV ends charging, rather than just considering the instantaneous power of the MG at the time when the charging application is submitted.

[0135] Step 3: Select the best charging station

[0136] In order to take into account the dual needs of user cost and MMG stability, the entropy weight method is used in this embodiment to integrate the user's time cost and the impact of EV charging on MMG to achieve the purpose of joint optimization.

[0137] Specifically, in this embodiment, an optimization model is defined, which includes multiple weighting coefficients, which reflect the relative importance between user cost and MMG impact. By adjusting these weighting coefficients, the relationship between user cost and MMG load can be flexibly weighed to ensure that the stability of MMG is improved while reducing user waiting time. The entropy weight method is an objective weighting method based on information entropy, which is used to determine the weight of each indicator in the multi-indicator evaluation. Assume that the nth car is in the The user time cost at the charging station is The average load of each MG is recorded as and , after normalization, they are recorded as and Then the entropy weight method is used to calculate and Weight and , thus combining the two into one parameter, the formula is as follows

[0138]

[0139] in is the total index of the nth vehicle at the jth charging station, , is the weight coefficient. In MTC-SLBMS, EV selects the charging station with the best comprehensive time cost and spatial load balancing index among the candidate solutions. In this strategy, the vehicle dispatch center selects a suitable charging station for the i-th electric vehicle according to the following rules :

[0140]

[0141]

[0142]

[0143] MTC-SLBMS combines the dual considerations of time cost and spatial load balancing, which is conducive to taking into account both the charging experience of EV users and the stability of MMG. and , a balance can be found between meeting the time needs of EV users and MMG load balancing.

[0144] Furthermore, the effect of the above technical solution is verified through simulation:

[0145] The initial load used in the simulation (i.e., the load consumption when the EV is not charged) is derived from actual data on electricity demand in California

[33] . After being scaled down, the data is randomly assigned to charging stations. The sampling interval of the data is 5 minutes. In the simulation, the total number of time samples is = 288, and the sampling time window is 0:00 - 23:55. All N EVs send charging requests in chronological order within this time window. We explore the average time cost of users and the load balancing effect under different matching strategies. The number of charging piles C in each charging station is set to 50, and the total number of vehicles N ranges from 150 to 1500.

[0146] Specifically, in this embodiment, a 34-node traffic network is taken as an example to verify the feasibility of the new allocation strategy. Figure 1 As shown, in an actual traffic network in a 20 km × 20 km city, the area contains 34 road network nodes and 54 road segments, and the area is divided into 7 MGs, each of which contains a charging station.

[0147] By average user time Cost is used to measure the time spent by users under different strategies. is defined as follows:

[0148]

[0149] in, represents the average time cost of users, Indicates The time the vehicle spends on the road, Indicates The waiting time of a vehicle in queue, Indicates The charging time of each vehicle is the total number of vehicles. In order to measure the load balance between MMGs at time t, the valley-peak ratio is used. as a performance indicator.

[0150] in, is defined as follows:

[0151]

[0152] The larger the value is, the smaller the valley-peak gap between MMGs at time t is, which is beneficial to the safe operation of the power grid. On this basis, the average valley-peak ratio is defined as To measure the load balance between MMGs in the entire sampling window, The definition is as follows:

[0153]

[0154] It should be noted that, in this embodiment, the average time cost of users and the load balancing effect under different matching strategies are explored, and the number C of charging piles in each charging station is set to 50. Figure 2 The changes of user time cost with the increase of the number of vehicles under different strategies are compared. Figure 3 The changes of average valley-to-peak ratio with the increase of vehicle number under different strategies are compared.

[0155] from Figure 2 and Figure 3 In the analysis, it was observed that the MTC-SLBMS algorithm showed significant effects in reducing the average time cost of users and improving the average valley-peak ratio. Specifically, the TMMS algorithm does provide the lowest average time cost for users under ideal conditions, but its main drawback is the unevenness of load distribution. In addition, although the ILBMS algorithm performs best in balancing the load of the power grid, it sacrifices the time cost of users, resulting in a significant increase in the average time cost of users. The experimental data show that the MTC-SLBMS algorithm not only reduces the total waiting time of users in the system, but also effectively improves the spatial load balancing of the power grid. This balancing strategy makes the MTC-SLBMS algorithm more flexible and adaptable in practical applications, and can dynamically adjust resource allocation according to user needs and the actual operating conditions of the power grid. The implementation of this algorithm helps to alleviate the pressure of the power grid during peak hours and reduce energy waste through more reasonable load distribution, thereby supporting the sustainable development of the power grid.

[0156] The embodiments of the present invention are described in detail above with reference to the accompanying drawings, but the present invention is not limited to the described embodiments. For those skilled in the art, various changes, modifications, substitutions and variations of these embodiments including components are made without departing from the principles and spirit of the present invention, and still fall within the scope of protection of the present invention.

Claims

1. A lightweight 3D human pose estimation method for uncalibrated multi-view cameras, characterized in that: The following steps are involved: Step S1, collect a 3D human pose estimation dataset of any number of viewpoints and define A 2D pose sequence of a dynamic perspective and a corresponding 3D pose sequence; Step S2, using the bidirectional Mamba model, extracting features of the multi-view two-dimensional posture sequence to obtain a prediction result; Step S3, pre-training and fine-tuning the bidirectional Mamba module; Step S4, using the multi-view dataset shot by an uncalibrated movable camera to train the bidirectional Mamba model end-to-end, output the three-dimensional posture of each view, align the postures of different views to one of the views through procrustes, and complete the human body posture estimation.

2. The lightweight 3D human pose estimation method for an uncalibrated multi-view camera according to claim 1, characterized in that: The step S1 specifically includes: collecting a 3D human posture estimation dataset of any number of viewpoints, using an uncalibrated movable camera; defining 2D pose sequence of dynamic viewpoints and the corresponding 3D pose sequence ,in is the number of viewing angles, is the length of the pose sequence, is the number of human skeleton joints defined by the model, is the number of input channels, is the number of output channels.

3. The lightweight 3D human pose estimation method for an uncalibrated multi-view camera according to claim 2, characterized in that: The bidirectional Mamba model comprises a preprocessing module and a plurality of groups of bidirectional Mamba modules connected in series; The preprocessing module takes the input 2D pose sequence Mapping into latent space ,in is the dimension of the latent space; Input multiple sets of bidirectional Mamba modules in series; each bidirectional Mamba module contains two branches, one of which first passes through the time Mamba module and then the space Mamba module, and the other first passes through the space Mamba module and then the time Mamba module. The two branches are fused by direct vector addition; after passing through multiple sets of bidirectional Mamba modules with the same model parameter structure, the prediction results are obtained. With the data set Calculate the loss and optimize the bidirectional Mamba model parameters by back-propagation.

4. The lightweight 3D human pose estimation method for an uncalibrated multi-view camera according to claim 3, characterized in that: The specific implementation process of the time Mamba module is as follows: For the input feature sequence Apply the normalization operation to normalize the feature sequence Mapped to two new feature spaces through two fully connected layers, respectively, and ; After mapping, the sequence Perform a 1D convolution and apply the SiLU activation function to get ; For the convolutional features Perform two linear transformations respectively to obtain new features and ; then After passing through the fully connected layer, the log-SoftPlus function is used to activate the output of the linear transformation to ensure that the intermediate parameters of the output time scale is a positive number; right and , and The two sets of parameters are respectively Sum and get the extended features and ,in is a learnable parameter; The transformed features , ,and As a Mamba module parameter, the feature As input to the Mamba module, get new features ; The feature and Add element by element, apply the fully connected layer and SiLU activation function, and get the output of the temporal Mamba module .

5. The lightweight 3D human pose estimation method for an uncalibrated multi-view camera according to claim 4, characterized in that: The specific implementation process of the spatial Mamba module is as follows: Output of the Mamba module at a given time , first transpose it along the last two dimensions, and get ,Will As the input of the spatial Mamba module, the subsequent calculation steps are the same as those of the temporal Mamba module, and we get .

6. The lightweight 3D human pose estimation method for an uncalibrated multi-view camera according to claim 5, characterized in that: In the bidirectional Mamba module, the output comes from the spatial Mamba module and time Mamba module , the output of the bidirectional Mamba module is obtained through adaptive fusion ; The specific process of adaptive fusion is as follows: Convert to , and then with Splicing ; right Use linear transformation to calculate attention and perform softmax operation to get attention weight ; then Split by the last dimension and ; Finally, the attention weights are used to and Perform weighted fusion to obtain : ； in Represents element-wise multiplication; The output of N bidirectional Mamba modules After a fully connected layer, the prediction result is obtained , used for the calculation of subsequent losses.

7. The lightweight 3D human pose estimation method for an uncalibrated multi-view camera according to claim 6, characterized in that: The specific implementation process of step S3 is as follows: Step S3.1: In the pre-training stage, orthogonal projection is used to project the prediction results into a complete 2D skeleton pose sequence, and then the partially defective 2D pose is obtained by random masking. , to achieve the purpose of data enhancement; in this stage, the 3D key point loss, speed loss and 2D reprojection loss are calculated as follows: ； in is the 3D keypoint loss, is the speed loss, is the 2D reprojection loss; and Represents the predicted and actual values ​​of the three-dimensional true pose; and Respectively represent the predicted value and actual value of the speed of the key point between video frames, namely: ; and Respectively represent the predicted value and actual value of the two-dimensional reprojection posture. Represents each video frame, Represents each key point, is the confidence of the 2D pose provided by the dataset; Step S3.2, in the fine-tuning phase, a task head is added to the pre-trained model to estimate the 3D pose; The 2D posture sequences of multiple perspectives in the same scene are superimposed and input into the bidirectional Mamba model. Then, some key points and perspectives are masked by random masking to simulate the occlusion challenges in the real environment. The random masking is implemented by adding a control gate in the input channel of each perspective. , masking joint points or viewing angles.

8. The lightweight 3D human pose estimation method for an uncalibrated multi-view camera according to claim 7, characterized in that: The specific implementation process of step S4 is as follows: a bidirectional Mamba model is trained end-to-end using a multi-view dataset shot by an uncalibrated mobile camera. The model outputs the three-dimensional posture of each view. The postures of different views are aligned to one of the views by translation, rotation and scaling through procrustes, so that the postures are aligned and the coordinate system is unified. In addition to calculating the three-dimensional key point loss, speed loss and two-dimensional reprojection loss, this stage also calculates the multi-view loss, which is as follows: First, the model predicts the 3D posture of each view output, which is recorded as: , ; Then use the first perspective as the reference view. If there are other perspectives, that is, , align the 3D pose estimated from other views with the pose estimated from the first view through the procrustes method; The posture after procrustes alignment is recorded as: ； Finally, procrustes alignment is used to calculate the multi-view consistency loss: ； The total loss at this stage is: ； in is a binary indicator function that determines whether to include multi-view consistency loss. represents the weighted multi-view consistency loss, , Represents the weight of the corresponding loss.

Images