Charging pile intelligent management method and system

Abstract

The invention belongs to the technical field of charging pile management, and discloses a charging pile intelligent management method and system. Comprising the following steps: carrying out data acquisition on a target area and charging piles needing to be managed in the target area to obtain corresponding charging pile data and area load data; uploading the data to a pre-constructed scheduling block chain; performing electric pile diagnosis on the corresponding charging pile based on the collected electric pile data to obtain a corresponding fault diagnosis result, judging whether the charging pile has a fault based on the fault diagnosis result, and if not, marking the corresponding charging pile as a healthy electric pile; if the fault exists, fault information is generated and fault early warning is carried out; performing power distribution load management on the corresponding healthy electric piles in combination with the collected regional load data; according to the invention, intelligent management of the charging pile is realized, the operation efficiency and safety of the charging pile are improved, and more efficient power distribution load management is realized in combination with the charging demand of a user.

Description

Technical Field

[0001] The present invention relates to the technical field of charging pile management, and more specifically, to a charging pile intelligent management method and system. Background Art

[0002] With the shortage of energy and the need for environmental protection, electric vehicles are widely promoted and applied to reduce the energy consumption of fuel vehicles and reduce the impact of exhaust emissions on the environment. With the accelerated development of new energy electric vehicles, the number of electric vehicles has increased sharply. The construction of charging stations has solved part of the charging problem of electric vehicles; however, the management of charging stations has become increasingly difficult;

[0003] Compared with the existing technology, the traditional intelligent management method of charging piles usually relies on manual operation, and the number of existing charging pile clusters is increasing, and the management efficiency of the corresponding charging piles is getting lower and lower, which not only reduces the charging efficiency, but also due to the lack of effective scheduling and management mechanism, the use of charging piles often cannot be monitored and adjusted in real time, resulting in the failure of charging piles and waste of load resources cannot be discovered in time.

[0004] In view of this, the present invention proposes a charging pile intelligent management method and system to solve the above problems. Summary of the invention

[0005] In order to overcome the above-mentioned defects of the prior art and to achieve the above-mentioned purpose, the present invention provides the following technical solutions:

[0006] A charging pile intelligent management method, comprising:

[0007] Step 1: Collect data on the target area and the charging piles that need to be managed in the target area, obtain the corresponding charging pile data and regional load data, and upload them to the pre-built scheduling blockchain;

[0008] Step 2: Performing electric pile diagnosis on the corresponding charging pile based on the collected electric pile data to obtain corresponding fault diagnosis results, and judging whether the charging pile has a fault based on the fault diagnosis results. If there is no fault, the corresponding charging pile is marked as a healthy electric pile; if there is a fault, fault information is generated and a fault warning is issued;

[0009] Step 3: Perform distribution load management on the corresponding healthy charging piles based on the collected regional load data.

[0010] Furthermore, data is collected on the target area and the charging piles to be managed in the target area to obtain corresponding charging pile data and regional load data; and the process of uploading the data to the pre-built scheduling blockchain includes:

[0011] Obtain the distribution topology structure corresponding to the charging piles to be managed in the target area, and divide the target area into regions based on the distribution topology structure to obtain a number of charging pile regions;

[0012] Setting a supervision cycle, and collecting data on charging piles in a corresponding charging pile area during a corresponding supervision cycle to obtain corresponding charging pile data, wherein the charging pile data includes the working status, charging data, and electricity price of the charging pile; wherein the charging data includes the charging pile temperature, charging voltage, and charging current;

[0013] At the same time, the regional load data corresponding to the corresponding charging pile area in the corresponding supervision period is obtained, and the regional load data includes the total load supply in the corresponding supervision period and the single load consumption corresponding to each charging pile;

[0014] Furthermore, the collected regional load data and charging pile data are preprocessed, and the preprocessed charging pile data and regional load data are uploaded to the pre-built scheduling blockchain.

[0015] Furthermore, the retrieved blockchain includes a master node, ordinary nodes and working nodes. The master node is used to manage transaction information between working nodes; the working node is used to add, delete or modify the information stored in the corresponding ordinary node, and is also used to store the corresponding regional load data; the ordinary node is used to store the uploaded charging pile data.

[0016] Furthermore, the process of dividing the target area into regions based on the distribution topology structure and obtaining several charging pile regions includes:

[0017] Read the obtained distribution topology structure, and mark the corresponding charging piles in the corresponding distribution topology structure as management nodes, and mark the distribution center to which the corresponding target area belongs as a distribution node;

[0018] The management node n generates a random value zn, zn∈(0,1); and compares it with the node threshold Tn;

[0019] If zn≤Tn, no other operations are performed; if zn>Tn, the corresponding management node n is marked as a regional node; where n=1,2,……,N1,N1>0 and N1 is an integer, and N1 represents the total number of management nodes;

[0020] The acquisition process based on the regional nodes traverses all management nodes and marks the management nodes that are not marked as regional nodes as allocation nodes;

[0021] Respectively obtain the position information of the distribution node, the regional node and the distribution node, wherein the position information refers to the coordinate position of the corresponding distribution node, the regional node and the distribution node in the world coordinate system;

[0022] Then, the trust factor XR between the corresponding regional node i and the assigned node j is obtained (i,j) ; Wherein, i=1, 2, ..., N2, N2 is a natural number, N2 represents the total number of regional nodes, j=1, 2, ..., N3, N3 represents the total number of allocated nodes, N3 is a natural number, and N2+N3=N1, N2<K;

[0023] The obtained trust factor is compared with the corresponding adaptive trust threshold; and the corresponding regional nodes and allocation nodes are divided according to the comparison result to obtain several charging areas;

[0024] Furthermore, the charging pile corresponding to the regional node transmits a cluster notification to other charging piles in the corresponding charging area, and obtains the network throughput and transmission delay between the corresponding charging piles during the corresponding transmission process, and compares them with the pre-set throughput threshold and delay threshold. If the network throughput and transmission delay both meet the corresponding throughput threshold and delay threshold, the corresponding charging area is marked as a charging pile area. If at least one of the network throughput and transmission delay does not meet the corresponding throughput threshold and delay threshold, the area is re-divided based on the acquisition process of the charging area to obtain a new charging area, and so on.

[0025] Furthermore, get the node threshold: In the formula, p represents the probability of management node n being selected as a regional node, r represents the number of rounds of selecting regional nodes; mod represents the modulus operation;

[0026] Trust Factor In the formula, XR (i,j) represents the trust factor between regional node i and allocation node j; lt and wt represent the length and width of the corresponding management area respectively; τ and μ represent trust metric parameters, which are determined in advance by technical personnel in this field according to actual needs; D (j,gl) represents the distance between the distribution node j and the distribution node gl;

[0027] In the formula, x gl 、x i 、x j Respectively represent the horizontal coordinates in the position information corresponding to the corresponding distribution node gl, management node i and management node j; D (i,gl) represents the distance between regional node i and distribution node gl; θ (i,j) represents the acute angle between the regional node i with the distribution node gl as the vertex and the distribution node j;

[0028] The formula for obtaining the adaptive trust threshold is:

[0029] Where D max represents the maximum distance between the management node and the corresponding distribution center; λ is the weight coefficient, In the formula, λ min and k are specific parameters, b i represents the total number of other management nodes adjacent to the corresponding regional node i; P i Represents the adaptive trust threshold corresponding to regional node i.

[0030] Furthermore, the process of performing electric pile diagnosis on the corresponding charging pile based on the collected electric pile data and obtaining the corresponding fault diagnosis result includes:

[0031] Read the charging data corresponding to each charging pile in the corresponding charging pile area;

[0032] Construct a two-dimensional rectangular coordinate system with time as the horizontal axis and charging data as the vertical axis, and map the corresponding charging data into the corresponding two-dimensional rectangular coordinates to obtain a corresponding charging data curve;

[0033] Based on the charging data curve, the charging data corresponding to several historical monitoring cycles are acquired, and based on the charging data curve, the charging prediction data corresponding to a future supervision cycle is predicted;

[0034] The supervision cycle is divided into intervals to obtain a plurality of time intervals; based on the charging data curve and the charging prediction data, the electric pile diagnostic data corresponding to the time interval to which the charging pile belongs at the current moment and the future Q time intervals are obtained, wherein the electric pile diagnostic data is composed of the real-time charging data of the corresponding charging pile and the charging prediction data corresponding to the Q time intervals; Q is a constant;

[0035] Input the charging pile diagnostic data into a pre-built fault supervision model to obtain corresponding fault supervision results, wherein the fault supervision results include the fault probability of the corresponding charging pile under different fault types in the current time interval and the next Q time intervals;

[0036] Set a probability threshold and compare the fault probability in the obtained fault supervision results with the corresponding probability threshold. If the corresponding fault probability is less than the probability threshold, the corresponding charging pile will be marked as a healthy charging pile; if the fault probability is not less than the corresponding probability threshold, indicating that the corresponding charging pile has the corresponding fault type, the corresponding charging pile will be marked as a faulty charging pile, and the fault information will be fed back to the corresponding staff.

[0037] Furthermore, the process of building a fault supervision model includes:

[0038] Obtain known fault types of the charging pile, and obtain several groups of historical fault data corresponding to several groups of corresponding fault types based on a big data algorithm; the historical fault data is the charging data corresponding to the charging pile under the corresponding fault type;

[0039] Obtain the fault data signal corresponding to the corresponding historical fault data; and decompose the obtained fault data signal C times based on the wavelet threshold function to obtain the corresponding high-frequency signal and low-frequency signal; C is a fixed constant;

[0040] Obtain the wavelet coefficient corresponding to the high-frequency signal in the cth decomposition layer, and use it as the fault feature vector corresponding to the corresponding fault data signal; where c∈C;

[0041] Count all fault feature vectors corresponding to the historical fault data of the corresponding fault type to obtain the corresponding fault feature set;

[0042] Obtain the fault feature set corresponding to all fault types and construct corresponding training data based on it;

[0043] The backbone network of the fault supervision model is defined as a BP neural network; the BP neural network includes an input layer, a hidden layer and an output layer, the input vector is input into the input layer of the corresponding BP neural network, and forward propagation is performed according to the input layer, the hidden layer and the output layer, and the output vector of the corresponding output layer is obtained;

[0044] Compare the output vector of the corresponding output layer with the expected output vector, and obtain the corresponding output error E;

[0045] The output error is compared with the expected error. If the output error meets the expected error, the training ends. If the output error does not meet the expected error, reverse propagation is performed according to the input layer, hidden layer and output layer, and the weights of neurons in the input layer, hidden layer and output layer are adjusted;

[0046] After adjusting the weights of each layer, the input vector is propagated in the positive direction again, and the "positive propagation" and "reverse propagation" processes are repeated and alternating, and the weights of each layer are continuously readjusted until the output error between the actual output vector and the expected output vector of the BP neural network meets the set expected error requirement. The iteration ends, and the weights of each connection are fixed; the corresponding model parameters are saved, and the corresponding fault supervision model is obtained;

[0047] Furthermore, the output vector of the output layer In the formula, S k represents the kth output vector corresponding to the output layer, k = 1, 2, ..., K, K > 0 and K is an integer, K represents the total number of output vectors corresponding to the output layer; w ykrepresents the initial weight w between the yth output vector of the hidden layer and the kth output vector of the corresponding output layer;

[0048] H y represents the y-th output vector corresponding to the hidden layer; y=1,2,……,Y,Y>0 and Y is an integer; Y represents the total number of output vectors corresponding to the hidden layer; In the formula, v py represents the initial weight v between the p-th input vector and the y-th output vector in the hidden layer; H P represents the pth input vector; f represents the activation function;

[0049] Output Error Where, d k Represents the expected output vector corresponding to the output vector of the kth output layer;

[0050] The corresponding weight adjustment formula is:

[0051]

[0052] In the formula, β is a constant, which is used to represent the learning efficiency of the corresponding BP neural network;

[0053] Δw yk represents the adjustment value of the initial weight w between the yth output vector of the hidden layer and the kth output vector of the corresponding output layer;

[0054] Δv py represents the adjustment value of the initial weight v between the p-th input vector and the y-th output vector in the hidden layer;

[0055] In the formula, represents partial differential calculation; δ_k represents the output error signal corresponding to the k-th output vector of the output layer.

[0056] Furthermore, the process of managing the distribution load of the corresponding healthy electric piles in combination with the collected regional load data includes:

[0057] Obtain the electric pile area described in the corresponding healthy electric pile, and obtain the regional load data corresponding to the corresponding electric pile area based on the scheduling blockchain;

[0058] Based on the regional load data, the historical single load consumption corresponding to several historical supervision cycles corresponding to the corresponding healthy charging pile is obtained;

[0059] At the same time, based on the regional load data, the real-time single load consumption corresponding to the real-time time interval to which the current moment belongs is obtained;

[0060] Based on the historical single load consumption, load forecasting is performed to obtain the expected load consumption corresponding to the current real-time time interval, and the deviation between it and the real-time single load consumption is calculated to obtain the corresponding load deviation;

[0061] Setting a load threshold, and comparing the obtained load deviation with the corresponding load threshold;

[0062] If the corresponding load deviation is not less than the load threshold, the corresponding load deviation is marked as a standby load;

[0063] If the corresponding load deviation is less than the load threshold, the ordinary node corresponding to the corresponding healthy charging pile generates corresponding replenishment transaction information; and feeds it back to the working node to which it belongs. The working node adjusts the charging pile strategy based on the received replenishment transaction information.

[0064] Furthermore, it is necessary to further explain that, in the specific implementation process, the process of adjusting the charging pile strategy based on the received replenishment transaction information includes:

[0065] The working node broadcasts the corresponding replenishment transaction information to other common nodes; after receiving the corresponding replenishment transaction information, the other common nodes obtain the corresponding reserve load surplus and feed it back to the working node; the working node obtains the corresponding reserve load total amount according to the received reserve load surplus and compares it with the deficit load amount in the corresponding replenishment transaction information;

[0066] If the total amount of standby load is higher than the amount of deficit load, the working node constructs a corresponding load transaction in combination with the replenishment transaction information; and based on the load transaction, load is transmitted from other ordinary nodes whose reserve load margin is not zero to the ordinary node that feeds back the replenishment transaction information; after the transmission is completed, the load transaction is broadcast to all ordinary nodes for storage;

[0067] If the total amount of standby load is not higher than the amount of the deficit load, the working node feeds back the adjustment transaction to the corresponding ordinary node;

[0068] The working node reduces the load of the corresponding healthy charging pile within the corresponding time interval based on the adjustment transaction, and adjusts the charging plan of the charging vehicle based on the load deviation of the charging piles before and after it, and then controls the corresponding healthy charging pile to supply power to the corresponding charging vehicle with the adjusted charging plan.

[0069] Furthermore, a charging pile intelligent management system includes:

[0070] Data collection module: used to collect data on the target area and the charging piles that need to be managed in the target area, obtain the corresponding charging pile data and regional load data, and upload them to the pre-built scheduling blockchain;

[0071] Fault diagnosis module; used to perform electric pile diagnosis on the corresponding charging pile based on the collected electric pile data, obtain the corresponding fault diagnosis result, and judge whether the charging pile has a fault based on the fault diagnosis result. If there is no fault, the corresponding charging pile is marked as a healthy electric pile; if there is a fault, fault information is generated and a fault warning is issued;

[0072] Load management module: used to manage the distribution load of the corresponding healthy charging piles in combination with the collected regional load data.

[0073] The technical effects and advantages of the charging pile intelligent management method and system of the present invention are as follows:

[0074] 1. It can obtain the working status, charging data, electricity price and other related parameters of the charging pile in real time, and upload these data to the scheduling blockchain. This not only enables managers to grasp the operation status of the charging pile at any time, but also improves the timeliness and accuracy of data processing, thereby effectively improving the management efficiency of the charging pile.

[0075] 2. Based on the distribution topology structure, the charging piles in the target area are divided into regions to obtain the corresponding charging pile areas. This helps to optimize the configuration of power resources according to the load demand and distribution of charging piles in each region, ensure the stable power supply of charging piles in each region, and improve the fault response speed and maintenance efficiency. It helps to balance the load of each region to avoid excessive or low load in some areas, resulting in waste of resources or insufficient power supply. At the same time, it reduces the data processing volume of the charging piles and improves the response efficiency of the charging piles. BRIEF DESCRIPTION OF THE DRAWINGS

[0076] Figure 1 A schematic diagram of a charging pile intelligent management method of the present invention;

[0077] Figure 2 This is a schematic diagram of a charging pile intelligent management system of the present invention. DETAILED DESCRIPTION

[0078] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only 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.

[0079] Example 1

[0080] See also Figure 1 As shown, the charging pile intelligent management method described in this embodiment includes:

[0081] Step 1: Collect data on the target area and the charging piles that need to be managed in the target area, obtain the corresponding charging pile data and regional load data, and upload them to the pre-built scheduling blockchain;

[0082] Step 2: Performing electric pile diagnosis on the corresponding charging pile based on the collected electric pile data to obtain corresponding fault diagnosis results, and judging whether the charging pile has a fault based on the fault diagnosis results. If there is no fault, the corresponding charging pile is marked as a healthy electric pile; if there is a fault, fault information is generated and a fault warning is issued;

[0083] Step 3: Perform distribution load management on the corresponding healthy charging piles in combination with the collected regional load data;

[0084] It should be further explained that, in the specific implementation process, data is collected on the target area and the charging piles to be managed in the target area, and the corresponding charging pile data and regional load data are obtained; and the data is uploaded to the pre-built scheduling blockchain, and the process includes:

[0085] Obtaining a distribution topology structure corresponding to the charging piles to be managed in the target area, and dividing the corresponding target area into regions based on the distribution topology structure to obtain a plurality of charging pile regions, wherein the charging pile region contains at least one charging pile to be managed; wherein the distribution topology structure refers to the distribution status of the charging piles to be managed in the corresponding target area and the connection status between the charging piles; and can be obtained based on the laying plan of the charging piles in the target area;

[0086] Setting a supervision cycle, and collecting data on the charging piles in the corresponding charging pile area during the corresponding supervision cycle to obtain corresponding charging pile data, wherein the charging pile data includes relevant parameters such as the working status of the charging pile, charging data, and electricity price; wherein the charging data includes the charging pile temperature, charging voltage, and charging current;

[0087] At the same time, the regional load data corresponding to the corresponding charging pile area in the corresponding supervision period is obtained, and the regional load data includes the total load supply in the corresponding supervision period and the single load consumption corresponding to each charging pile; wherein the total load supply is determined by the total distribution load and the total basic load, and the total distribution load refers to the load that can be provided by the distribution center to which the corresponding charging pile area belongs according to the pre-established distribution plan; the total basic load refers to the load required by other electrical equipment in the corresponding charging pile area except the charging piles;

[0088] Furthermore, the collected regional load data and electric pile data are preprocessed, and the preprocessed electric pile data and regional load data are uploaded to the pre-built scheduling blockchain; the retrieved blockchain includes a master node, a common node and a working node, the master node is used to manage transaction information between working nodes; the working node is used to add, delete or modify the information stored in the corresponding common node, and is also used to store the corresponding regional load data; the common node is used to store the uploaded electric pile data,

[0089] Among them, a master node is connected to several working nodes, and the number of working nodes is determined by the number of charging pile areas corresponding to the corresponding target area; a working node is connected to several common nodes, and one working node corresponds to one charging pile area, and the number of common nodes connected to the working node is determined by the number of charging piles required to be managed in the charging pile area to which it belongs;

[0090] It should be further explained that, in the specific implementation process, the data preprocessing includes data cleaning, data conversion and data normalization; data cleaning is mainly to correct and fill in the errors, missing and inconsistent data in the original data to ensure the integrity of the data; data conversion is to unify the data of different formats and units into standard formats and units to facilitate subsequent data analysis and processing; data normalization is to normalize the data to eliminate the dimension and scale differences between the data to make the data comparable;

[0091] It should be further explained that, in the specific implementation process, the process of dividing the target area based on the distribution topology structure and obtaining several charging pile areas includes:

[0092] Read the obtained distribution topology structure, and mark the corresponding charging piles in the corresponding distribution topology structure as management nodes, and mark the distribution center to which the corresponding target area belongs as a distribution node;

[0093] The management node n generates a random value zn, zn∈(0,1); and compares it with the node threshold Tn. If zn≤Tn, no other operation is performed; if zn>Tn, the corresponding management node n is marked as a regional node; wherein n=1,2,……,N1,N1>0 and N1 is an integer, N1 represents the total number of management nodes; wherein the formula for obtaining the node threshold Tn is: Where p represents the probability of management node n being selected as a regional node, r represents the number of rounds for selecting regional nodes, and the value of r is determined by the total number of regional nodes selected. For example, if the total number of regional nodes selected in the current round is not less than the constant K, secondary marking is performed from the marked regional nodes until the total number of selected regional nodes is less than the constant K; mod represents the modulo operation;

[0094] The acquisition process based on the regional nodes traverses all management nodes and marks the management nodes that are not marked as regional nodes as allocation nodes;

[0095] Respectively obtain the position information of the distribution node, the regional node and the distribution node, wherein the position information refers to the coordinate position of the corresponding distribution node, the regional node and the distribution node in the world coordinate system;

[0096] Then, the trust factor XR between the corresponding regional node i and the assigned node j is obtained (i,j) ; Wherein, i=1, 2, ..., N2, N2 is a natural number, N2 represents the total number of regional nodes, j=1, 2, ..., N3, N3 represents the total number of allocated nodes, N3 is a natural number, and N2+N3=N1, N2<K;

[0097] The trust factor In the formula, XR (i,j) represents the trust factor between regional node i and allocation node j; lt and wt represent the length and width of the corresponding management area respectively; τ and μ represent trust metric parameters, which are determined in advance by technical personnel in this field according to actual needs; D (j,gl) represents the distance between the distribution node j and the distribution node gl; θ (i,j) It represents the acute angle between the regional node i with the distribution node gl as the vertex and the distribution node j. The corresponding mathematical formula is as follows:

[0098] In the formula, x gl 、x i 、x j Respectively represent the horizontal coordinates in the position information corresponding to the corresponding distribution node gl, management node i and management node j; D (i,gl) represents the distance between regional node i and distribution node gl;

[0099] The obtained trust factor is compared with the corresponding adaptive trust threshold; wherein the formula for obtaining the adaptive trust threshold is: P i =2pλ×exp{(-λ)×[(-k)log(1+k× Where D max represents the maximum distance between the management node and the corresponding distribution center; λ is the weight coefficient, In the formula, λ min and k are specific parameters, depending on actual needs, b i represents the total number of other management nodes adjacent to the corresponding regional node i; P i represents the adaptive trust threshold corresponding to regional node i;

[0100] If the corresponding trust factor is less than or equal to the adaptive trust threshold, it means that the regional node i and the allocation node j are not mutually trusted nodes, and no other operations are performed;

[0101] If the corresponding trust factor is greater than the adaptive trust threshold, it indicates that the regional node i and the allocation node j are mutually trusted nodes, and the corresponding regional node i and the management node j are divided into the same sub-area; repeat until the corresponding target area is divided into several charging areas;

[0102] Furthermore, the charging pile corresponding to the regional node transmits cluster notifications to other charging piles in the corresponding charging area, obtains the network throughput and transmission delay between the corresponding charging piles during the corresponding transmission process, and compares them with the respective pre-set throughput thresholds and delay thresholds. If both the network throughput and the transmission delay meet the corresponding throughput thresholds and delay thresholds, the corresponding charging area is marked as a charging pile area. If at least one of the network throughput and the transmission delay does not meet the corresponding throughput threshold and delay threshold, the area is re-divided based on the acquisition process of the charging area to obtain a new charging area, and so on. Among them, the rule for judging whether the corresponding network throughput and transmission delay meet the corresponding throughput threshold and delay threshold is: if the corresponding network throughput (or transmission delay -) is higher than the corresponding throughput threshold (or delay threshold), it is determined that the corresponding network throughput (or transmission delay) meets the corresponding throughput threshold (or delay threshold);

[0103] It should be further explained that, in the specific implementation process, the process of performing electric pile diagnosis on the corresponding charging pile based on the collected electric pile data and obtaining the corresponding fault diagnosis result includes:

[0104] Taking any charging pile area as an example, read the charging data corresponding to each charging pile in the corresponding charging pile area;

[0105] Constructing a two-dimensional rectangular coordinate system with time as the horizontal axis and charging data as the vertical axis, and mapping the corresponding charging data into the corresponding two-dimensional rectangular coordinates to obtain a corresponding charging data curve; the charging data curve includes a temperature curve, a voltage curve and a current curve;

[0106] Based on the charging data curve, the charging data corresponding to several historical monitoring cycles are acquired, and based on the charging data curve, the charging prediction data corresponding to a future supervision cycle is predicted;

[0107] The supervision cycle is divided into intervals to obtain a plurality of time intervals; based on the charging data curve and the charging prediction data, the electric pile diagnostic data corresponding to the time interval to which the charging pile belongs at the current moment and the future Q time intervals are obtained, wherein the electric pile diagnostic data is composed of the real-time charging data of the corresponding charging pile and the charging prediction data corresponding to the Q time intervals; Q is a constant;

[0108] Input the charging pile diagnostic data into a pre-built fault supervision model to obtain corresponding fault supervision results, wherein the fault supervision results include the fault probability of the corresponding charging pile under different fault types in the current time interval and the next Q time intervals;

[0109] A probability threshold is set, and the fault probability in the obtained fault supervision result is compared with the corresponding probability threshold. If the corresponding fault probability is less than the probability threshold, it indicates that the corresponding charging pile does not have the corresponding fault type, then the corresponding charging pile is marked as a healthy charging pile, and the ordinary node corresponding to the corresponding charging pile is set as a public node; if the fault probability is not less than the corresponding probability threshold, it indicates that the corresponding charging pile has the corresponding fault type, then the corresponding charging pile is marked as a faulty charging pile, and the fault information is fed back to the corresponding staff, and the fault information includes fault warning, fault type and location information of the corresponding charging pile; at the same time, the ordinary node corresponding to the corresponding faulty charging pile is set as a private node, that is, it does not participate in the scheduling of the charging pile until the staff feedbacks the maintenance completion notification, cancels the faulty charging pile mark, and changes the corresponding private node to a public node;

[0110] It should be further explained that, in the specific implementation process, the construction process of the fault supervision model includes:

[0111] Obtain known fault types of the charging pile, and obtain several groups of historical fault data corresponding to several groups of corresponding fault types based on a big data algorithm; the historical fault data is the charging data corresponding to the charging pile under the corresponding fault type;

[0112] Taking any fault type as an example, the fault data signal corresponding to the corresponding historical fault data is obtained; and the obtained fault data signal is decomposed C times based on the wavelet threshold function to obtain the corresponding high-frequency signal and low-frequency signal; wherein C represents the number of decomposition layers corresponding to the decomposition process of the corresponding wavelet threshold function; C is a fixed constant;

[0113] Obtain the wavelet coefficient corresponding to the high-frequency signal in the cth decomposition layer, and use it as the fault feature vector corresponding to the corresponding fault data signal; where c∈C; (i.e., one fault data signal can correspond to multiple fault feature vectors, and the specific number of fault feature vectors depends on the number of decomposition layers;)

[0114] All fault feature vectors corresponding to the historical fault data of the corresponding fault type are counted to obtain the corresponding fault feature set; wherein, in the corresponding statistical process, the fault feature vectors counted will be repeatedly checked, and if a certain fault feature vector appears multiple times, the repeated fault feature vector will be eliminated;

[0115] Obtain the fault feature set corresponding to all fault types and construct corresponding training data based on it;

[0116] The backbone network of the fault supervision model is defined as a BP neural network, which is used to learn the nonlinear mapping relationship between the input vector and the corresponding fault type; wherein the input vector of the BP neural network is the fault feature vector in the training data;

[0117] The BP neural network includes an input layer, a hidden layer and an output layer. The input vector is input into the input layer of the corresponding BP neural network, and forward propagation is performed according to the input layer, the hidden layer and the output layer, and the output vector of the corresponding output layer is obtained. In the formula, S k represents the kth output vector corresponding to the output layer, k = 1, 2, ..., K, K > 0 and K is an integer, K represents the total number of output vectors corresponding to the output layer; w yk represents the initial weight w between the yth output vector of the hidden layer and the kth output vector of the corresponding output layer;

[0118] H y represents the y-th output vector corresponding to the hidden layer; y=1,2,……,Y,Y>0 and Y is an integer; Y represents the total number of output vectors corresponding to the hidden layer; In the formula, v py represents the initial weight v between the p-th input vector and the y-th output vector in the hidden layer; H P represents the pth input vector of the input;

[0119] f represents an activation function, which is selected in advance by those skilled in the art according to actual needs; wherein the initial weight w and the initial weight v are both network parameters of the corresponding BP neural network, which are set by the staff according to actual needs when constructing the corresponding BP neural network;

[0120] Then, the output vector of the corresponding output layer is compared with the expected output vector, and the corresponding output error E is obtained; the formula for obtaining the corresponding output error E is:

[0121] Where, d k Represents the expected output vector corresponding to the output vector of the kth output layer;

[0122] The output error is compared with the expected error. If the output error meets the expected error, the training ends. If the output error does not meet the expected error, the input layer, hidden layer and output layer are propagated in the reverse direction, and the weights of the neurons in the input layer, hidden layer and output layer are adjusted. The formula for the corresponding weight adjustment is:

[0123]

[0124] In the formula, β is a constant, which is used to represent the learning efficiency of the corresponding BP neural network;

[0125] Δw yk represents the adjustment value of the initial weight w between the yth output vector of the hidden layer and the kth output vector of the corresponding output layer;

[0126] Δv py represents the adjustment value of the initial weight v between the p-th input vector and the y-th output vector in the hidden layer;

[0127] δ_k represents the output error signal corresponding to the k-th output vector of the output layer; the corresponding acquisition formula is as follows: In the formula, represents partial differential calculation;

[0128] When the weights of each layer are adjusted, the input vector is propagated in the positive direction again, and the "input forward propagation" and "output reverse propagation" processes are repeatedly alternated to continuously readjust the weights of each layer until the output error between the actual output vector and the expected output vector of the BP neural network meets the set expected error requirement. The iteration is then terminated and the weights of each connection are fixed; the corresponding model parameters are saved to obtain the corresponding fault supervision model; the expected error and the expected output vector are determined according to actual needs.

[0129] It should be further explained that, in the specific implementation process, the process of managing the distribution load of the corresponding healthy electric piles in combination with the collected regional load data includes:

[0130] Obtain the electric pile area described in the corresponding healthy electric pile, and obtain the regional load data corresponding to the corresponding electric pile area based on the scheduling blockchain;

[0131] Based on the regional load data, the historical single load consumption corresponding to several historical supervision cycles corresponding to the corresponding healthy charging pile is obtained;

[0132] At the same time, based on the regional load data, the real-time single load consumption corresponding to the real-time time interval to which the current moment belongs is obtained;

[0133] Based on the historical single load consumption, load forecasting is performed to obtain the expected load consumption corresponding to the current real-time time interval, and the deviation between it and the real-time single load consumption is calculated to obtain the corresponding load deviation;

[0134] Setting a load threshold, and comparing the obtained load deviation with the corresponding load threshold;

[0135] If the corresponding load deviation is not less than the load threshold, it indicates that the expected load consumption of the corresponding healthy charging pile is higher than the real-time single load consumption (i.e., there is a load surplus in the corresponding healthy charging pile), and the corresponding load deviation is marked as a standby load;

[0136] If the corresponding load deviation is less than the load threshold, it indicates that the expected load consumption of the corresponding healthy electric pile is lower than the real-time single load consumption (i.e., the corresponding healthy electric pile has a load deficit); then the ordinary node corresponding to the corresponding healthy electric pile generates corresponding replenishment transaction information; and feeds it back to the working node to which it belongs, and the working node adjusts the charging pile strategy based on the received replenishment transaction information;

[0137] It should be further explained that, in the specific implementation process, the process of adjusting the charging pile strategy based on the received replenishment transaction information includes:

[0138] Then, the working node broadcasts the corresponding replenishment transaction information to other common nodes; when other common nodes receive the corresponding replenishment transaction information, they obtain the corresponding reserve load margin and feed it back to the working node; wherein, if the common node itself has a load deficit, the corresponding reserve load margin is zero;

[0139] The working node obtains the corresponding total amount of standby load according to the received surplus of standby load, and compares it with the deficit load in the corresponding replenishment transaction information;

[0140] If the total amount of standby load is higher than the amount of deficit load, the working node constructs a corresponding load transaction in combination with the replenishment transaction information; and based on the load transaction, load is transmitted from other ordinary nodes whose reserve load margin is not zero to the ordinary node that feeds back the replenishment transaction information; after the transmission is completed, the load transaction is broadcast to all ordinary nodes for storage;

[0141] If the total amount of standby load is not higher than the amount of the deficit load, the working node feeds back the adjustment transaction to the corresponding ordinary node;

[0142] Furthermore, the working node reduces the load of the corresponding healthy charging pile within the corresponding time interval based on the adjustment transaction, and adjusts the charging plan of the charging vehicle based on the load deviation of the charging piles before and after it, and then controls the corresponding healthy charging pile to supply power to the corresponding charging vehicle with the adjusted charging plan, and feeds back the adjusted charging plan to the corresponding car owner; wherein the specific adjustment process depends on the actual load deficit, and at the same time, it must be constrained that the adjusted expected charging time cannot exceed the original expected charging time.

[0143] The present invention divides the charging piles in the target area into regions based on the distributed topological structure, obtains corresponding charging pile regions, and uploads the data collected from each charging pile region to the scheduling blockchain, which helps to grasp the operation status of the charging piles at any time, and also improves the timeliness and accuracy of data processing, thereby effectively improving the management efficiency of the charging piles; by constructing a fault supervision model, not only can the charging interruption phenomenon caused by charging pile failure be reduced, but also the service life of the charging pile can be extended and the maintenance cost can be reduced; distribution load management of healthy charging piles can ensure the load balance of the charging piles, and can also improve the utilization efficiency of power resources and reduce energy waste.

[0144] Example 2

[0145] See also Figure 2 As shown, the part not described in detail in this embodiment is described in Example 1, which provides a charging pile intelligent management system, including:

[0146] Data collection module: used to collect data on the target area and the charging piles that need to be managed in the target area, obtain the corresponding charging pile data and regional load data, and upload them to the pre-built scheduling blockchain;

[0147] Fault diagnosis module; used to perform electric pile diagnosis on the corresponding charging pile based on the collected electric pile data, obtain the corresponding fault diagnosis result, and judge whether the charging pile has a fault based on the fault diagnosis result. If there is no fault, the corresponding charging pile is marked as a healthy electric pile; if there is a fault, fault information is generated and a fault warning is issued;

[0148] Load management module: used to manage the distribution load of the corresponding healthy electric piles in combination with the collected regional load data;

[0149] The modules are connected to each other via wired and / or wireless means to achieve data transmission between modules.

[0150] Example 3

[0151] This embodiment discloses an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the operation mode of the charging pile intelligent management method and system provided above is implemented.

[0152] Since the electronic device introduced in this embodiment is an electronic device used to implement a charging pile intelligent management method and system in the embodiment of this application, based on the charging pile intelligent management method and system introduced in the embodiment of this application, the technical personnel of this field can understand the specific implementation of the electronic device of this embodiment and its various variations, so how the electronic device implements the method in the embodiment of this application is not described in detail here. As long as the technical personnel of this field implement the electronic device used in the charging pile intelligent management method and system in the embodiment of this application, it belongs to the scope of protection of this application.

[0153] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters and thresholds in the formula are set by technicians in this field according to actual conditions.

[0154] The above is only a preferred embodiment of the present invention, and the protection scope of the present invention is not limited to the above embodiments. All technical solutions under the concept of the present invention belong to the protection scope of the present invention. It should be pointed out that for ordinary technical users in this technical field, some improvements and modifications without departing from the principle of the present invention should also be regarded as the protection scope of the present invention.

Claims

1. A charging pile intelligent management method, characterized in that: include: Step 1: Collect data on the target area and the charging piles that need to be managed in the target area to obtain corresponding charging pile data and regional load data; And upload it to the pre-built scheduling blockchain; Step 2: Performing an electric pile diagnosis on the corresponding charging pile based on the collected electric pile data to obtain a corresponding fault diagnosis result, and judging whether the charging pile has a fault based on the fault diagnosis result. If there is no fault, marking the corresponding charging pile as a healthy electric pile; If there is a fault, fault information is generated and a fault warning is issued; Step 3: Perform distribution load management on the corresponding healthy charging piles based on the collected regional load data.

2. A charging pile intelligent management method according to claim 1, characterized in that: Collect data on the target area and the charging piles that need to be managed in the target area to obtain corresponding charging pile data and regional load data; The process of uploading it to the pre-built dispatch blockchain includes: Obtain the distribution topology structure corresponding to the charging piles to be managed in the target area, and divide the target area into regions based on the distribution topology structure to obtain a number of charging pile regions; Setting a supervision cycle, and collecting data on charging piles in a corresponding charging pile area during a corresponding supervision cycle to obtain corresponding charging pile data, wherein the charging pile data includes the working status, charging data, and electricity price of the charging pile; wherein the charging data includes the charging pile temperature, charging voltage, and charging current; At the same time, the regional load data corresponding to the corresponding charging pile area in the corresponding supervision period is obtained, and the regional load data includes the total load supply in the corresponding supervision period and the single load consumption corresponding to each charging pile; Then, the collected regional load data and electric pile data are preprocessed, and the preprocessed electric pile data and regional load data are uploaded to the pre-built scheduling blockchain; The retrieved blockchain includes a master node, an ordinary node and a working node. The master node is used to manage the working node; the working node is used to add, delete or modify the information stored in the corresponding ordinary node, and is also used to store the corresponding regional load data; the ordinary node is used to store the uploaded electric pile data.

3. A charging pile intelligent management method according to claim 2, characterized in that: The process of dividing the target area into regions based on the distribution topology and obtaining several charging pile regions includes: Read the obtained distribution topology structure, and mark the corresponding charging piles in the corresponding distribution topology structure as management nodes, and mark the distribution center to which the corresponding target area belongs as a distribution node; The management node n generates a random value zn, zn∈(0,1); and compares it with the node threshold Tn; If zn≤Tn, no other operations are performed; if zn>Tn, the corresponding management node n is marked as a regional node; where n=1,2,……,N1,N1>0 and N1 is an integer, and N1 represents the total number of management nodes; The acquisition process based on the regional nodes traverses all management nodes and marks the management nodes that are not marked as regional nodes as allocation nodes; Respectively obtain the position information of the distribution node, the regional node and the distribution node, wherein the position information refers to the coordinate position of the corresponding distribution node, the regional node and the distribution node in the world coordinate system; Then, the trust factor XR between the corresponding regional node i and the assigned node j is obtained (i,j) ; Wherein, i=1, 2, ..., N2, N2 is a natural number, N2 represents the total number of regional nodes, j=1, 2, ..., N3, N3 represents the total number of allocation nodes, N3 is a natural number; the obtained trust factor is compared with the corresponding adaptive trust threshold; and the corresponding regional nodes and allocation nodes are divided according to the comparison result to obtain several charging areas; The charging pile corresponding to the regional node transmits a cluster notification to other charging piles in the corresponding charging area, and obtains the network throughput and transmission delay between the corresponding charging piles during the corresponding transmission process, and compares them with the respective pre-set throughput thresholds and delay thresholds. If the network throughput and transmission delay both meet the corresponding throughput threshold and delay threshold, the corresponding charging area is marked as a charging pile area. If at least one of the network throughput and transmission delay does not meet the corresponding throughput threshold and delay threshold, the area is re-divided based on the acquisition process of the charging area to obtain a new charging area, and so on.

4. A charging pile intelligent management method according to claim 3, characterized in that: The node threshold In the formula, p represents the probability of management node n being selected as a regional node, r represents the number of rounds of selecting regional nodes; mod represents the modulus operation; Trust Factor In the formula, XR (i,j) represents the trust factor between regional node i and allocation node j; lt and wt represent the length and width of the corresponding management area respectively; τ and μ represent trust measurement parameters; D (j,gl) represents the distance between the distribution node j and the distribution node gl; In the formula, x gl 、x i 、x j Respectively represent the horizontal coordinates in the position information corresponding to the corresponding distribution node gl, management node i and management node j; D (i,gl) represents the distance between regional node i and distribution node gl; θ (i,j) represents the acute angle between the regional node i with the distribution node gl as the vertex and the distribution node j; The formula for the adaptive trust threshold is: Where D max represents the maximum distance between the management node and the corresponding distribution center; λ is the weight coefficient, In the formula, λ min and k are specific parameters, b i represents the total number of other management nodes adjacent to the corresponding regional node i; P i Represents the adaptive trust threshold corresponding to regional node i.

5. A charging pile intelligent management method according to claim 2, characterized in that: The process of performing electric pile diagnosis on the corresponding charging pile based on the collected electric pile data and obtaining the corresponding fault diagnosis result includes: Read the charging data corresponding to each charging pile in the corresponding charging pile area; Construct a two-dimensional rectangular coordinate system with time as the horizontal axis and charging data as the vertical axis, and map the corresponding charging data into the corresponding two-dimensional rectangular coordinates to obtain a corresponding charging data curve; Based on the charging data curve, the charging data corresponding to several historical monitoring cycles are acquired, and based on the charging data curve, the charging prediction data corresponding to a future supervision cycle is predicted; The supervision cycle is divided into intervals to obtain a plurality of time intervals; based on the charging data curve and the charging prediction data, the electric pile diagnostic data corresponding to the time interval to which the charging pile belongs at the current moment and the future Q time intervals are obtained, wherein the electric pile diagnostic data is composed of the real-time charging data of the corresponding charging pile and the charging prediction data corresponding to the Q time intervals; Q is a constant; Input the charging pile diagnostic data into a pre-built fault supervision model to obtain corresponding fault supervision results, wherein the fault supervision results include the fault probability of the corresponding charging pile under different fault types in the current time interval and the next Q time intervals; Set a probability threshold and compare the fault probability in the obtained fault supervision results with the corresponding probability threshold. If the corresponding fault probability is less than the probability threshold, the corresponding charging pile will be marked as a healthy charging pile; if the fault probability is not less than the corresponding probability threshold, the corresponding charging pile will be marked as a faulty charging pile, and the fault information will be fed back to the corresponding staff.

6. A charging pile intelligent management method according to claim 5, characterized in that: The construction process of the fault supervision model includes: Obtain known fault types of the charging pile, and obtain several groups of historical fault data corresponding to several groups of corresponding fault types based on a big data algorithm; the historical fault data is the charging data corresponding to the charging pile under the corresponding fault type; Obtain the fault data signal corresponding to the corresponding historical fault data; and decompose the obtained fault data signal C times based on the wavelet threshold function to obtain the corresponding high-frequency signal and low-frequency signal; C is a fixed constant; Obtain the wavelet coefficient corresponding to the high-frequency signal in the cth decomposition layer, and use it as the fault feature vector corresponding to the corresponding fault data signal; where c∈C; Count all fault feature vectors corresponding to the historical fault data of the corresponding fault type to obtain the corresponding fault feature set; Obtain the fault feature set corresponding to all fault types and construct corresponding training data based on it; The backbone network of the fault supervision model is defined as a BP neural network; the BP neural network includes an input layer, a hidden layer and an output layer, the input vector is input into the input layer of the corresponding BP neural network, and forward propagation is performed according to the input layer, the hidden layer and the output layer, and the output vector of the corresponding output layer is obtained; Compare the output vector of the corresponding output layer with the expected output vector, and obtain the corresponding output error E; The output error is compared with the expected error. If the output error meets the expected error, the training ends. If the output error does not meet the expected error, reverse propagation is performed according to the input layer, hidden layer and output layer, and the weights of the neurons in the input layer, hidden layer and output layer are adjusted; When the weights of each layer are adjusted, the input vector is propagated in the positive direction again, and the forward and reverse propagation processes are repeated and alternating, and the weights of each layer are continuously readjusted until the output error between the actual output vector and the expected output vector of the BP neural network meets the set expected error requirement. The iteration is ended, and the weights of each connection are fixed; the corresponding model parameters are saved to obtain the corresponding fault supervision model.

7. A charging pile intelligent management method according to claim 6, characterized in that: The output vector of the output layer In the formula, S k represents the kth output vector corresponding to the output layer, k = 1, 2, ..., K, K > 0 and K is an integer, K represents the total number of output vectors corresponding to the output layer; w yk represents the initial weight w between the yth output vector of the hidden layer and the kth output vector of the corresponding output layer; H y represents the y-th output vector corresponding to the hidden layer; y=1,2,……,Y,Y>0 and Y is an integer; Y represents the total number of output vectors corresponding to the hidden layer; In the formula, v py represents the initial weight v between the p-th input vector and the y-th output vector in the hidden layer; H P represents the pth input vector; f represents the activation function; Output Error Where, d k Represents the expected output vector corresponding to the output vector of the kth output layer; The corresponding weight adjustment formula is: In the formula, β is a constant, which is used to represent the learning efficiency of the corresponding BP neural network; Δw yk represents the adjustment value of the initial weight w between the yth output vector of the hidden layer and the kth output vector of the corresponding output layer; Δv py represents the adjustment value of the initial weight v between the p-th input vector and the y-th output vector in the hidden layer; In the formula, represents partial differential calculation; δ_k represents the output error signal corresponding to the k-th output vector of the output layer.

8. The charging pile intelligent management method according to claim 5, characterized in that: The process of managing the distribution load of the corresponding healthy charging piles in combination with the collected regional load data includes: Obtain the electric pile area described in the corresponding healthy electric pile, and obtain the regional load data corresponding to the corresponding electric pile area based on the scheduling blockchain; Based on the regional load data, the historical single load consumption corresponding to several historical supervision cycles corresponding to the corresponding healthy charging pile is obtained; At the same time, based on the regional load data, the real-time single load consumption corresponding to the real-time time interval to which the current moment belongs is obtained; Based on the historical single load consumption, load forecasting is performed to obtain the expected load consumption corresponding to the current real-time time interval, and the deviation between it and the real-time single load consumption is calculated to obtain the corresponding load deviation; Setting a load threshold, and comparing the obtained load deviation with the corresponding load threshold; If the corresponding load deviation is not less than the load threshold, the corresponding load deviation is marked as a standby load; If the corresponding load deviation is less than the load threshold, the ordinary node corresponding to the corresponding healthy charging pile generates corresponding replenishment transaction information; and feeds it back to the working node to which it belongs. The working node adjusts the charging pile strategy based on the received replenishment transaction information.

9. A charging pile intelligent management method according to claim 8, characterized in that: The process of adjusting the charging pile strategy based on the received replenishment transaction information includes: The working node broadcasts the corresponding replenishment transaction information to other common nodes; after receiving the corresponding replenishment transaction information, the other common nodes obtain the corresponding reserve load surplus and feed it back to the working node; the working node obtains the corresponding reserve load total amount according to the received reserve load surplus and compares it with the deficit load amount in the corresponding replenishment transaction information; If the total amount of standby load is higher than the amount of deficit load, the working node constructs a corresponding load transaction in combination with the replenishment transaction information; and based on the load transaction, load is transmitted from other ordinary nodes whose reserve load margin is not zero to the ordinary node that feeds back the replenishment transaction information; after the transmission is completed, the load transaction is broadcast to all ordinary nodes for storage; If the total amount of standby load is not higher than the amount of the deficit load, the working node feeds back the adjustment transaction to the corresponding ordinary node; The working node reduces the load of the corresponding healthy charging pile within the corresponding time interval based on the adjustment transaction, and adjusts the charging plan of the charging vehicle based on the load deviation of the charging piles before and after it, and then controls the corresponding healthy charging pile to supply power to the corresponding charging vehicle with the adjusted charging plan.

10. A charging pile intelligent management system, characterized in that: A charging pile intelligent management method for implementing any one of claims 1 to 9, comprising: Data collection module: used to collect data on the target area and the charging piles that need to be managed in the target area, obtain the corresponding charging pile data and regional load data, and upload them to the pre-built scheduling blockchain; Fault diagnosis module; used to perform electric pile diagnosis on the corresponding charging pile based on the collected electric pile data, obtain the corresponding fault diagnosis result, and judge whether the charging pile has a fault based on the fault diagnosis result. If there is no fault, the corresponding charging pile is marked as a healthy electric pile; if there is a fault, fault information is generated and a fault warning is issued; Load management module: used to manage the distribution load of the corresponding healthy charging piles in combination with the collected regional load data.

Images