Low-carbon park cluster topology prediction and timing method and system based on power line carrier

Through BP neural network prediction topological relationships and synchronous group game algorithm optimization path selection, the topological perception and time delay problems of power line carrier timing method in low-carbon smart park groups are solved, and high-precision, low-delay time synchronization is achieved, and low-delay time synchronization is supported. Time synchronization with wide cost coverage.

CN115694560BActive Publication Date: 2024-11-19NORTH CHINA ELECTRIC POWER UNIV +1
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202211277609.4
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-10-19
Publication Date
2024-11-19
Estimated Expiration
2042-10-19

AI Technical Summary

Technical Problem

The prior art power line carrier timing method in low-carbon smart park groups has poor topological perception capabilities, large delay in timing packets, large errors, and difficult to optimize trade-offs between multiple performance indicators, resulting in unstable time synchronization.

Method used

The low-carbon park group topology prediction and timing method based on power line carrier is adopted, and the terminal topology adjacency relationship is monitored using BP neural network prediction, and the timing packet path selection is optimized through the synchronous group game algorithm to achieve high-precision and low-latency time synchronization.

Benefits of technology

It realizes high-precision, low-delay, wide coverage time synchronization of monitoring terminals and electrical equipment in the group of low-carbon smart parks, reduces costs, adapts to changes in complex topological structures, and improves timing performance.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN115694560B_ABST
    Figure CN115694560B_ABST
Patent Text Reader

Abstract

The present invention discloses a method and system for predicting and timing a low-carbon park cluster topology based on power line carriers. The method uses power line carriers to collect characteristic data between low-carbon park topologies, historical timing delays and error data of power line carrier timing data packets, and accurately predicts the topological adjacency relationship of monitoring terminals in a low-carbon smart park cluster through a back propagation (BP) neural network. Based on the topological adjacency relationship of the monitoring terminals, a synchronous group game is used to optimize the path selection decision of the timing data packet, and further time synchronization is provided to the monitoring terminals and electrical equipment inside the park through power line carriers, so as to achieve high-precision, low-latency and wide-coverage time synchronization of the low-carbon smart park cluster.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention belongs to the field of power line carrier communication, and in particular relates to a low-carbon park group topology prediction and timing method based on power line carrier. Background Art

[0002] At present, my country is in the stage of energy transformation. Under the background of tight power supply and the goal of "carbon peak and carbon neutrality", building a low-carbon smart park group with renewable energy as the main body is of great significance for energy conservation, emission reduction and energy efficiency improvement. The low-carbon operation of the smart park group is achieved by deploying monitoring terminals on electrical equipment such as photovoltaic panels, wind turbines, and electric vehicle charging piles in the park to monitor the power flow and equipment operation of the park, relying on flexible load control, carbon footprint monitoring, power spot market transactions and other power services. In this process, the information collection of the monitoring terminal and the stable operation of the power business require precise time synchronization. On the one hand, the collection of low-carbon park operation data such as voltage, current, and phase angle, which are strongly related to time, needs to meet the consistency of collection time to ensure the accuracy of line fault distance measurement, dynamic monitoring of phase and power angle, and verification of unit and power grid parameters. On the other hand, various power services also need to be carried out on the same time scale. Time asynchrony will lead to inconsistent business response times of various energy entities in the park group, resulting in out-of-step oscillation and asynchronous operation of the low-carbon smart park group, which will affect the overall safety of the park group. The existing Beidou satellite timing method requires the installation of satellite positioning modules and satellite receiving antennas on each monitoring terminal and electrical equipment, which is expensive and inconvenient to construct. In addition, satellite timing is easily affected by external factors, and the stability of time synchronization is poor. The timing accuracy of 5G and optical fiber timing methods is high, but the transmission distance of 5G is short. In order to achieve wide-area time synchronization, a large number of high-power 5G base stations need to be deployed. The high cost of optical fiber construction is also not conducive to the time synchronization of large-scale low-carbon smart park clusters. Therefore, there is an urgent need for a low-cost, wide-coverage time synchronization method that supports the safe and stable operation of low-carbon smart park clusters.

[0003] Power line carrier can achieve low-cost and wide-coverage time synchronization. Power line carrier uses existing power lines to transmit timing information. There is no need to set up an additional network. It has low deployment cost and wide coverage. Monitoring terminals and power equipment can achieve time synchronization by connecting to power lines. The time synchronization method of low-carbon smart park groups using power line carrier is divided into two steps. First, the low-carbon smart park gateway collects accurate timing information through satellite, 5G, optical fiber and other clock sources to achieve time synchronization between different gateways in the park group. Secondly, the park gateway sends timing signals through power line carriers to achieve wide-coverage time synchronization of monitoring terminals and electrical equipment within the park. However, the current power line carrier timing method faces problems such as poor park topology perception, large delay of timing data packets, and large timing errors. The specific introduction is as follows:

[0004] First, the frequent switching of electrical equipment in low-carbon smart park clusters leads to complex and changeable topology of monitoring terminals. The existing topology prediction method sends path search signals through the gateway and analyzes the feedback path information to perform topology prediction. In the case of complex park environment and frequent changes in park topology, there are problems such as long path search time and poor topology prediction accuracy, which leads to large timing delay and high overhead.

[0005] Second, the timing accuracy of the power line carrier data packets sent by the low-carbon smart park is different, and the channel characteristics such as electromagnetic interference and path gain between the sender and the receiver change dynamically, which makes it difficult to optimize the path selection of the timing data packets. In addition, since there is no available path or the path conditions are poor between some data packet receivers and the timing data packet senders, multi-hop transmission is required to obtain the data packet, which increases the complexity of the timing data packet path selection optimization, resulting in large timing delays and errors.

[0006] Third, the indicators for measuring timing performance include timing delay, timing error, and data packet transmission reliability, and each performance indicator constrains each other. For example, the timing error of some timing data packet senders in the low-carbon smart park is small, but the channel conditions between the timing monitoring terminal to be synchronized are poor. Using this timing sender for timing may lead to problems such as large timing delay and high packet loss rate. Therefore, how to achieve a trade-off between multiple performance indicators is also a problem.

[0007] Technical solution of prior art 1:

[0008] Topology prediction method based on power line carrier;

[0009] Disadvantages of prior art 1:

[0010] The existing technology requires sending a large number of power line carrier signals to measure and monitor the path information between terminals, and analyzing the path information to predict the network topology, which leads to a large waste of network resources and a serious increase in topology prediction time. It is also difficult to adapt to the rapid changes in the topology structure of low-carbon smart park groups and cannot meet the real-time requirements of topology prediction.

[0011] Technical solution of the second prior art:

[0012] Time synchronization method based on power line carrier;

[0013] Disadvantages of the second prior art:

[0014] The second existing technology selects a transmitter for each receiver with the goal of maximizing the signal-to-interference-noise ratio to optimize the power line carrier timing path selection strategy. It does not consider the joint optimization of multiple indicators such as the differentiated timing error and timing delay of the transmitter, resulting in a serious reduction in the timing accuracy of the receiver and the overall timing performance, and is unable to ensure high-precision time synchronization in the complex environment of the low-carbon smart park group. Summary of the invention

[0015] The present invention aims to solve the technical problems existing in the background technology, and aims to provide a low-carbon park group topology prediction and timing method based on power line carrier, which uses power line carrier to collect the characteristic data between low-carbon park topologies, the historical timing delay and error data of power line carrier timing data packets, and accurately predicts the topological adjacency relationship of the monitoring terminals of the low-carbon smart park group through the back propagation (BP) neural network. Based on the topological adjacency relationship of the monitoring terminals, the synchronization group game is used to optimize the path selection decision of the timing data packet, and further the power line carrier is used to provide timing for the monitoring terminals and electrical equipment inside the park, so as to realize the high-precision, low-latency and wide-coverage time synchronization of the low-carbon smart park group.

[0016] In order to solve the technical problem, the technical solution of the present invention is:

[0017] A low-carbon park cluster topology prediction and timing method based on power line carrier, the method comprising:

[0018] By analyzing the actual distribution of smart gateways and monitoring terminals in the smart park cluster and the precise timing information data packets, a power line carrier timing system model is constructed;

[0019] Based on the power line carrier timing system model, under the constraint of signal-to-interference-noise ratio, the optimization problem is constructed by optimizing the path selection strategy of the power line carrier timing data packet.

[0020] Obtain the historical topology information and real-time monitoring data of power line carrier stored in the low-carbon smart park database, predict the topology of monitoring terminals, and obtain the topological adjacency of monitoring terminals;

[0021] Based on the signal-to-interference-to-noise ratio between monitoring terminals in the real-time monitoring data of power line carriers, the topological adjacency relationship of monitoring terminals and the constructed optimization problem, the synchronization group game algorithm is used to optimize the decision of the timing information data packet path to obtain the optimal timing decision result, which realizes the time synchronization of the park group monitoring terminals and electrical equipment.

[0022] Furthermore, the analysis of the actual distribution of smart gateways and monitoring terminals in the smart park group and the precise timing information data packets is specifically as follows:

[0023] Analyze the actual distribution of smart gateways and monitoring terminals in the smart park cluster, determine the set and express it as S = {s0, s1, ..., s i ,...,s I}, where s0 represents the smart gateway, s1, s i and I Respectively represent the 1st, i, and Ith monitoring terminals; if the connection topology of the monitoring terminals in the park is unknown, a total of R rounds of timing processes are considered, and the set is determined and expressed as R = {1, ..., r, ..., R}, where 1, r, and R represent the 1st, r, and R rounds of timing, respectively;

[0024] Analyze the precise timing information data packet to determine s i For the data packet sender, The packet receiving end; the packet path selection indicator variable is a binary variable It indicates that, Indicates the timing of the rth round From i Receive the data packet, otherwise Identify binary variables is the topological connection indicator variable between the intelligent gateway and each monitoring terminal. Indicates i and Connect, otherwise

[0025] Furthermore, the power line carrier timing system model includes: a power line carrier timing data packet transmission model, a timing delay model and a cumulative timing error model;

[0026] The power line carrier timing data packet transmission model is constructed as follows:

[0027] Considering the complex electromagnetic environment and noise interference of the low-carbon smart park, the transmission of power line carrier timing data packets is realized based on orthogonal frequency division multiplexing technology. From i The rate at which packets are received is:

[0028]

[0029] Among them, N s is the OFDM symbol transmission rate, is the transmission power, represents the path gain; where and σ0 are the path frequency response, electromagnetic interference and Gaussian white noise power respectively; represents the signal-to-interference-noise ratio gap, that is, the ability to resist signal-to-interference-noise ratio fading; where P e represents the target bit error rate, Q -1(x) indicates The inverse function of .

[0030] Construct the timing delay model; specifically:

[0031] Each round of timing is divided into several time slots of length τ. At the beginning of each time slot, s i Start connecting Transmit power line carrier timing data packets; in the multi-hop transmission of the low-carbon smart park monitoring terminal topology, the rth round of timing The delay of receiving a data packet is determined by the sender s i The delay of receiving data packets and From i The single-hop transmission delay of receiving a data packet consists of two parts, expressed as:

[0032]

[0033] Among them, among them, Indicates that in the rth round of timing, the node Receive from node s i The total delay of the data packet, Indicates i The number of time slots it takes to receive a packet, for From i The single-hop transmission delay of receiving data packets is expressed as:

[0034]

[0035] Among them, U r Indicates the size of the rth round of timing data packets.

[0036] Construct the cumulative timing error model, specifically:

[0037] Determine the rth round Cumulative timing error of received data packets for:

[0038]

[0039] in, The timing for the rth round From i The cumulative timing error of the received data packet is:

[0040]

[0041] in, Indicates the rth round of timing From iThe single-hop timing error caused by receiving data packets obeys the normal distribution with a mean error of α and a standard deviation of β, that is, The rth round of timing s i Cumulative timing error of received data packets.

[0042] Furthermore, the optimization goal is to minimize the weighted sum of the average timing delay and timing error per round by optimizing the path selection strategy of the power line carrier timing data packet under the constraint of the signal-to-interference-to-noise ratio. The optimization problem is specifically constructed as follows:

[0043]

[0044] Among them, χ is a weight parameter used to dynamically weigh the timing delay and timing error; C1 and C2 are path selection constraints, indicating the timing of each round of the receiving end. Can receive data packets from at most one sender; C3 means With s i When not connected, Unable to i Receive data packets; To ensure the transmission quality of the power line carrier timing data packets, a constraint is imposed on the signal-to-interference-to-noise ratio of the path, namely C4:

[0045]

[0046] Further, the monitoring terminal topology prediction specifically includes:

[0047] Select a timing path for precise timing information packets based on the topological adjacency of the monitoring terminals;

[0048] Input the historical topology information into the preset BP neural network for pre-training to obtain a trained BP neural network model;

[0049] Based on the constructed power line carrier timing system model, the real-time monitoring data of the power line carrier is input into the trained BP neural network model to obtain the topological adjacency relationship of the monitoring terminals.

[0050] Furthermore, the inputting of the historical topology information into a preset BP neural network for pre-training specifically includes:

[0051] The topological adjacency matrix of the park monitoring terminals, the characteristic data between topologies, the historical power line carrier timing delay and the historical power line carrier timing error stored in the low-carbon smart park database are input into the preset BP neural network for pre-training;

[0052] The preset BP neural network is constructed specifically as follows:

[0053] The BP neural network consists of an input layer, a hidden layer, and an output layer. The number of neurons in the input layer is X, the number of neurons in each hidden layer is Y, and the number of neurons in the output layer is Z.

[0054] Among them, the hidden layer parameters are Q(A,B), Q is the BP neural network parameter matrix, A is the hidden layer weight matrix, and B is the threshold matrix;

[0055] The input signal matrix of the BP neural network input layer suitable for monitoring terminal topology prediction is expressed as:

[0056]

[0057] in, Represent the input signals of the 1st, xth, and Xth input layer neurons respectively;

[0058] The activation function of the BP neural network is expressed as:

[0059]

[0060] Among them, κ is the activation function parameter; the hidden layer function of the BP neural network suitable for monitoring terminal topology prediction is:

[0061]

[0062] in, is the weight between the input layer neurons and the hidden layer neurons, b y is the hidden layer threshold, N y is the output of the yth neuron in the hidden layer;

[0063] The output layer function of the BP neural network suitable for monitoring terminal topology prediction is:

[0064]

[0065] in, is the weight between the hidden layer neurons and the output layer neurons, c z is the output layer threshold, G z is the output of the zth neuron;

[0066] The loss function of the BP neural network is expressed as:

[0067]

[0068] Among them, M is the number of samples, k is the output of the neural network training sample, and k *is the expected output, m is the dimension of the data; the weights are updated according to the negative gradient direction of the error; when the input signal matrix and the expected output are given, each input signal matrix is ​​iterated repeatedly, and when all the input signal matrix data are iterated repeatedly in turn, after the sample training is completed, it is determined whether the indicator function meets the accuracy; if the indicator function meets the accuracy, the training is stopped, otherwise, the training is repeated until the accuracy is met, and a trained BP neural network model suitable for monitoring terminal topology prediction is obtained.

[0069] Furthermore, the real-time monitoring data of the power line carrier is input into the trained BP neural network model to obtain the topological adjacency relationship of the monitoring terminal, which specifically includes:

[0070] At the beginning of each round of timing process, the real-time monitoring data of the power line carrier between the monitoring terminals is input into the trained BP neural network model to obtain the topological connection relationship of the monitoring terminals in this round of timing; the monitoring terminal topological adjacency relationship matrix W r It can be expressed by the following formula:

[0071]

[0072] Furthermore, before optimizing the decision of the timing information data packet path by using the synchronization group game algorithm, the method further includes:

[0073] First, in the rth round of timing, a power line carrier timing data packet transmitter and multiple timing data packet receivers are defined as a multicast group, denoted as The utility function can be expressed as follows:

[0074]

[0075] in, for The number of data packet receivers in

[0076] The synchronization group game of the rth round of timing is defined as Game r (Φ r ,V), where Φ represents the set of monitoring terminals participating in the game, satisfying V represents the value of the synchronization group, defined as:

[0077]

[0078] In each synchronization group, whether the synchronization group allows the data packet receiver to join depends on its contribution to the value of the synchronization group. If the receiver seriously increases the average timing delay and timing error of the multicast group, it will be rejected; therefore, the receiver in the timing synchronization group of the rth round Benefits is defined as the contribution to the sync group value, calculated as:

[0079]

[0080] in, express Join Sync Group

[0081] Furthermore, the synchronization group game algorithm is used to optimize the decision of the timing information data packet path, specifically:

[0082] Step 1: Synchronization group initialization:

[0083] According to the signal-to-interference-noise ratio and topological adjacency between the monitoring terminals, the priority of each receiving terminal to the senders of all received timing information packets is calculated, and the preference order is arranged in descending order of signal-to-interference-noise ratio; for the receiving terminal If there is a satisfying The sender of the at the same time, In Temporarily join a sync group where s i First in priority; if the receiving end Does not satisfy any sender The receiving end Does not participate in this round of timing;

[0084] Step 2: Create a sync group:

[0085] When all synchronization groups in this time slot are temporarily established, the receiving end, for example According to the formula The calculated benefit is the optimization goal to choose to join other synchronization groups; if but Select Leave current sync group Join the sync group The receiving end in can continuously choose to change the synchronization group according to the benefits until no receiving end chooses to change the synchronization group;

[0086] Step 3: Final synchronization group establishment and power line carrier timing data packet transmission:

[0087] Repeat step 2 until all receiving ends do not change their selections; the time when the synchronization group is finally formed is called the decision point; after this, the transmitter in each synchronization group starts to transmit the power line carrier timing data packet in the same synchronization group; when each transmitter in all synchronization groups completes the timing, the synchronization group of this time slot is disbanded;

[0088] Step 4: Next time slot transmission:

[0089] At the end of each time slot, check whether all synchronization groups have completed transmission; then, the receiving end that receives the power line carrier timing data packet in the current time slot will serve as the sending end for the next time slot transmission; the algorithm goes to step 1; when each monitoring terminal completes timing, this round of timing ends.

[0090] A low-carbon park cluster topology prediction and timing system based on power line carrier, the system is applied to the control layer, and the system includes:

[0091] The first building module is used to build a power line carrier timing system model by analyzing the actual distribution of smart gateways and monitoring terminals in the smart park group and the precise timing information data packets;

[0092] The second building module is used to construct an optimization problem by optimizing the path selection strategy of the power line carrier timing data packet based on the power line carrier timing system model under the signal-to-interference-noise ratio constraint;

[0093] The topology prediction module is used to obtain the historical topology information and real-time monitoring data of the power line carrier stored in the low-carbon smart park database, predict the topology of the monitoring terminal, and obtain the topological adjacency relationship of the monitoring terminal;

[0094] The decision optimization module is used to optimize the decision of the timing information data packet path based on the signal-to-interference-noise ratio between monitoring terminals in the real-time monitoring data of the power line carrier, the topological adjacency relationship of the monitoring terminals and the constructed optimization problem, and obtain the optimal timing decision result by using the synchronization group game algorithm. The decision result realizes the time synchronization of the park group monitoring terminals and electrical equipment.

[0095] Further, the first building module includes: an analysis unit and a first building unit;

[0096] The analysis unit is used to determine that there is an intelligent gateway and I monitoring terminals in the park, and determine the set and express it as S = {s0, s1, ..., s i ,...,s I}, where s0 represents the smart gateway, s1, s i and I Respectively represent the 1st, i, and Ith monitoring terminals; if the connection topology of the monitoring terminals in the park is unknown, a total of R rounds of timing processes are considered, and the set is determined and expressed as R = {1, ..., r, ..., R}, where 1, r, and R represent the 1st, r, and R rounds of timing, respectively;

[0097] Determine i For the data packet sender, The packet receiving end; the packet path selection indicator variable is a binary variable It indicates that, Indicates the timing of the rth round From i Receive the data packet, otherwise Identify binary variables is the topological connection indicator variable between the intelligent gateway and each monitoring terminal. Indicates i and Connect, otherwise

[0098] Further, the first construction unit is used to construct the power line carrier timing data packet transmission model, specifically:

[0099] Considering the complex electromagnetic environment and noise interference of the low-carbon smart park, the transmission of power line carrier timing data packets is realized based on orthogonal frequency division multiplexing technology. From i The rate at which packets are received is:

[0100]

[0101] Among them, N s is the OFDM symbol transmission rate, is the transmission power, represents the path gain; where and σ0 are the path frequency response, electromagnetic interference and Gaussian white noise power respectively; represents the signal-to-interference-noise ratio gap, that is, the ability to resist signal-to-interference-noise ratio fading; where P e represents the target bit error rate, Q -1 (x) indicates The inverse function of .

[0102] Used to construct the timing delay model; specifically:

[0103] Each round of timing is divided into several time slots of length τ. At the beginning of each time slot, s i Start connecting Transmit power line carrier timing data packets; in the multi-hop transmission of the low-carbon smart park monitoring terminal topology, the rth round of timing The delay of receiving a data packet is determined by the sender s i The delay of receiving data packets and From i The single-hop transmission delay of receiving a data packet consists of two parts, expressed as:

[0104]

[0105] in, Indicates that in the rth round of timing, the node Receive from node s iThe total delay of the data packet, Indicates i The number of time slots it takes to receive a packet, for From i The single-hop transmission delay of receiving data packets is expressed as:

[0106]

[0107] Among them, U r Indicates the size of the rth round of timing data packets.

[0108] Used to construct the cumulative timing error model, specifically:

[0109] Determine the rth round Cumulative timing error of received data packets for:

[0110]

[0111] in, The timing for the rth round From i The cumulative timing error of the received data packet is:

[0112]

[0113] in, Indicates the rth round of timing From i The single-hop timing error caused by receiving data packets obeys the normal distribution with a mean error of α and a standard deviation of β, that is, The rth round of timing s i Cumulative timing error of received data packets.

[0114] Furthermore, the second construction module includes: constructing an optimization problem unit;

[0115] The optimization problem unit is used to minimize the weighted sum of the average timing delay and timing error per round under the constraint of signal-to-interference-noise ratio by optimizing the path selection strategy of the power line carrier timing data packet. The optimization problem is specifically constructed as follows:

[0116]

[0117] Among them, χ is a weight parameter used to dynamically weigh the timing delay and timing error; C1 and C2 are path selection constraints, indicating the timing of each round of the receiving end. Can receive data packets from at most one sender; C3 means With s i When not connected, Unable to i Receive data packets; To ensure the transmission quality of the power line carrier timing data packets, a constraint is imposed on the signal-to-interference-to-noise ratio of the path, namely C4:

[0118]

[0119] Furthermore, the topology prediction module includes:

[0120] A monitoring unit, configured to select a timing path for a precise timing information data packet based on a monitoring terminal topology adjacency relationship;

[0121] A pre-training unit is used to input the historical topology information into a preset BP neural network for pre-training to obtain a trained BP neural network model;

[0122] The inputting of the historical topology information into a preset BP neural network for pre-training specifically includes:

[0123] The topological adjacency matrix of the park monitoring terminals, the characteristic data between topologies, the historical power line carrier timing delay and the historical power line carrier timing error stored in the low-carbon smart park database are input into the preset BP neural network for pre-training;

[0124] The preset BP neural network is constructed specifically as follows:

[0125] The BP neural network consists of an input layer, a hidden layer, and an output layer. The number of neurons in the input layer is X, the number of neurons in each hidden layer is Y, and the number of neurons in the output layer is Z.

[0126] Among them, the hidden layer parameters are Q(A,B), Q is the BP neural network parameter matrix, A is the hidden layer weight matrix, and B is the threshold matrix;

[0127] The input signal matrix of the BP neural network input layer suitable for monitoring terminal topology prediction is expressed as:

[0128]

[0129] in, Represent the input signals of the 1st, xth, and Xth input layer neurons respectively;

[0130] The activation function of the BP neural network is expressed as:

[0131]

[0132] Among them, κ is the activation function parameter; the hidden layer function of the BP neural network suitable for monitoring terminal topology prediction is:

[0133]

[0134] in, is the weight between the input layer neurons and the hidden layer neurons, b y is the hidden layer threshold, N y is the output of the yth neuron in the hidden layer;

[0135] The output layer function of the BP neural network suitable for monitoring terminal topology prediction is:

[0136]

[0137] in, is the weight between the hidden layer neurons and the output layer neurons, c z is the output layer threshold, G z is the output of the zth neuron;

[0138] The loss function of the BP neural network is expressed as:

[0139]

[0140] Among them, M is the number of samples, k is the output of the neural network training sample, and k * is the expected output, m is the dimension of the data; the weights are updated according to the negative gradient direction of the error; when the input signal matrix and the expected output are given, each input signal matrix is ​​iterated repeatedly, and when all input signal matrix data are iterated in turn, after the sample training is completed, it is determined whether the indicator function meets the accuracy; if the indicator function meets the accuracy, the training is stopped, otherwise the training is repeated until the accuracy is met, and a trained BP neural network model suitable for monitoring terminal topology prediction is obtained;

[0141] A solving unit is used to input the real-time monitoring data of the power line carrier into the trained BP neural network model based on the constructed power line carrier timing system model to obtain the topological adjacency relationship of the monitoring terminal;

[0142] The real-time monitoring data of the power line carrier is input into the trained BP neural network model to obtain the topological adjacency relationship of the monitoring terminal, which specifically includes:

[0143] At the beginning of each round of timing process, the real-time monitoring data of the power line carrier between the monitoring terminals is input into the trained BP neural network model to obtain the topological connection relationship of the monitoring terminals in this round of timing; the monitoring terminal topological adjacency relationship matrix W r It can be expressed by the following formula:

[0144]

[0145] Furthermore, the decision optimization module includes:

[0146] Synchronization group initialization unit: used to calculate the priority of each receiving end to the sender of all received timing information data packets according to the signal interference and noise ratio and topological adjacency between the monitoring terminals. The preference order is arranged in descending order of signal interference and noise ratio. If there is a satisfying The sender of the at the same time, In Temporarily join a sync group where s i First in priority; if the receiving end Does not satisfy any sender The receiving end Does not participate in this round of timing;

[0147] Establishing synchronization group unit: used when all synchronization groups of this time slot are temporarily established, the receiving end, for example According to the formula The calculated benefit is the optimization goal to choose to join other synchronization groups; if but Select Leave current sync group Join the sync group The receiving end in can continuously choose to change the synchronization group according to the benefits until no receiving end chooses to change the synchronization group;

[0148] Transmission establishment unit, used for final synchronization group establishment and power line carrier timing data packet transmission: specifically:

[0149] Repeat the processing method of establishing the synchronization group unit until all receiving ends do not change their selection; the time when the synchronization group is finally formed is called the decision point; after this, the transmitting end in each synchronization group starts to transmit the power line carrier timing data packet in the same synchronization group; when each transmitting end in all synchronization groups completes the timing, the synchronization group of this time slot is disbanded;

[0150] The detection unit is used to check whether all synchronization groups have completed transmission at the end of each time slot; then, the receiving end that receives the power line carrier timing data packet in the current time slot will serve as the sending end for the next time slot transmission; the algorithm is transferred to the synchronization group initialization unit; when each monitoring terminal completes the timing, this round of timing ends.

[0151] A computer-readable storage medium stores computer-executable instructions, which are used to implement any of the above methods when executed by a processor.

[0152] Compared with the prior art, the advantages of the present invention are:

[0153] 1) The present invention proposes a low-carbon smart park cluster topology prediction and timing system based on power line carrier, including a clock source layer, a gateway layer, a monitoring terminal layer and a control layer. Driven by the control layer, the power line carrier intelligent gateway of the low-carbon smart park cluster selects the timing information transmitted by Beidou satellite, 5G, and optical fiber time synchronization network to achieve time synchronization of the intelligent gateway of the low-carbon smart park cluster; secondly, the intelligent gateway executes the proposed high-precision, low-latency topology prediction and timing method based on power line carrier, perceives the park topology structure online, and optimizes the power line carrier timing data packet path to provide timing for the monitoring terminals and electrical equipment inside the park, achieving low-cost and wide-coverage time synchronization at the monitoring terminal layer of the low-carbon smart park cluster.

[0154] 2) The present invention establishes a path selection optimization model that weights and minimizes the timing delay and timing error based on the signal-to-interference-noise ratio constraint, ensures the reliability of timing data packet transmission through the signal-to-interference-noise ratio constraint, adjusts the optimization target weight according to the actual situation of the low-carbon smart park, and realizes the dynamic trade-off between the timing delay and timing error performance, so as to meet the timing needs of different low-carbon smart park groups, and support the park group's flexible load control, carbon footprint monitoring, electricity spot market transactions and other businesses to operate efficiently and stably on the same time scale.

[0155] 3) The present invention proposes a high-precision, low-latency topology prediction and timing method based on power line carrier. First, the historical topology information stored in the low-carbon smart park group database is imported, the mapping relationship between the monitoring terminal topology adjacency matrix and the input data is fitted, and the BP neural network model for topology prediction is trained. At the beginning of each round of timing process, the power line carrier monitoring data such as the signal-to-interference-noise ratio and electromagnetic interference between the monitoring terminals are input into the training model, and the topological adjacency relationship of the monitoring terminals is perceived online to achieve reliable, real-time, and low-cost prediction of the topological structure of the smart park monitoring terminal. Secondly, each round of timing is based on the predicted topological structure. Under the constraint of the signal-to-interference-noise ratio, the goal is to minimize the weighted sum of the average timing delay and timing error per round, so as to form a synchronous group to transmit the timing data packet, thereby ensuring high-precision, low-latency time synchronization in the complex environment of the low-carbon smart park group. BRIEF DESCRIPTION OF THE DRAWINGS

[0156] Figure 1 , low-carbon smart park cluster topology prediction and timing system based on power line carrier;

[0157] Figure 2 , Schematic diagram of the power line carrier sensing topology prediction method for low-carbon smart park clusters;

[0158] Figure 3 , Flowchart of the power line carrier timing method for low-carbon smart park groups based on synchronous group game. DETAILED DESCRIPTION

[0159] The specific implementation mode of the present invention is described below in conjunction with embodiments:

[0160] It should be noted that the structures, proportions, sizes, etc. shown in this specification are only used to match the contents disclosed in the specification so that people familiar with this technology can understand and read them, and are not used to limit the conditions under which the present invention can be implemented. Any structural modification, change in proportional relationship or adjustment of size should still fall within the scope of the technical content disclosed in the present invention without affecting the effects and purposes that can be achieved by the present invention.

[0161] At the same time, the terms such as "upper", "lower", "left", "right", "middle" and "one" cited in this specification are only for the convenience of description and are not used to limit the scope of implementation of the present invention. Changes or adjustments to their relative relationships should be regarded as the scope of implementation of the present invention without substantially changing the technical content.

[0162] Embodiment 1:

[0163] like Figure 1 As shown in the figure, the topology prediction and timing system for low-carbon smart park clusters includes a clock source layer, a gateway layer, a terminal layer, and a control layer. Among them, the clock source layer uses Beidou satellites, 5G, and optical fiber time synchronization networks to provide standard time information for the gateway layer. The gateway layer is composed of intelligent gateways of low-carbon smart park clusters. It performs local clock calibration by receiving timing information from the clock source layer, and uses multicast to send power line carrier timing data packets to the low-carbon smart park cluster monitoring terminals of the terminal layer to ensure the time synchronization between the monitoring terminals and electrical equipment within the park. The terminal layer deploys low-carbon smart park monitoring terminals on electrical equipment such as photovoltaic panels, wind turbines, and electric vehicle charging piles to provide the gateway layer with various types of electrical equipment information such as carbon emissions, voltage, current, active / reactive power, and charging and discharging power, and feeds back the timing error and timing delay of the timing data packet to the control layer to optimize the timing data packet path selection strategy. In addition, the low-carbon smart park monitoring terminal forwards the flexible load control, carbon footprint monitoring, power spot market transactions and other power business instructions issued by the load aggregator and the power grid dispatching center to control the distributed resources of the low-carbon smart park cluster. The control layer deploys functions such as topology prediction and power line timing data packet path selection optimization suitable for low-carbon smart park clusters, ensuring that the distributed resource regulation of the low-carbon smart park cluster is carried out on the same time scale.

[0164] The control layer interacts with each layer to complete the low-cost and wide-coverage time synchronization of the low-carbon smart park group. The control layer interacts with the clock source layer to dynamically adjust the broadcast frequency of Beidou satellite, 5G, and optical fiber timing signals based on the link quality of each clock source. The control layer interacts with the gateway layer and selects appropriate timing signals for the low-carbon smart park intelligent gateway in combination with the timing accuracy of each clock source in the clock source layer to obtain standard time information. The control layer interacts with the terminal layer to construct a low-carbon smart park group database, establish a terminal topology adjacency matrix based on a comprehensive analysis of a neural network, realize topology prediction, and optimize the path selection strategy of the power line carrier timing data packet based on the feedback timing delay and timing error based on the synchronization group game, realize the time synchronization of the park group monitoring terminal and electrical equipment, and ensure the consistent response of the distributed resource regulation of the low-carbon smart park group.

[0165] High-precision and low-latency topology prediction and timing method based on power line carrier

[0166] The present invention designs a high-precision, low-latency topology prediction and timing method based on power line carrier for low-carbon smart park clusters, which mainly includes four stages: power line carrier timing model construction, optimization problem modeling, low-carbon smart park cluster power line carrier perception topology prediction method, and low-carbon smart park cluster power line carrier timing method based on synchronous group game. The specific introduction is as follows:

[0167] (1) Construction of power line carrier timing model

[0168] The present invention considers the scenario where the intelligent gateway of a low-carbon smart park group obtains accurate timing information through satellite, 5G, and optical fiber time synchronization network, and sends power line carrier timing data packets to the monitoring terminals in the park for timing. Assume that there is an intelligent gateway and I monitoring terminals in the park, and the set is represented by S = {s0, s1, ..., s i ,...,s I}, where s0 represents the smart gateway, s1, s i and IRepresent the 1st, i, and Ith terminals respectively. Assume that the terminal connection topology of the park is unknown. Consider a total of R rounds of timing processes, and the set is expressed as R = {1,...,r,...,R}, where 1, r, and R represent the 1st, r, and R rounds of timing respectively. The intelligent gateway and the terminal communicate through the power line carrier. Each round of timing starts with the intelligent gateway sending a power line carrier timing data packet and ends when all terminals in the park receive the timing data packet. The length of each round of timing is equal to the time when the last terminal receives the timing data packet. In each round of timing, the intelligent gateway sends the timing data packet to the terminal. After the terminal that receives the timing data packet first completes time synchronization, it becomes a new sender and distributes the timing data packet to the terminal that has not received the data packet until all terminals receive the timing data packet. In particular, the intelligent gateway can only serve as a data packet sender.

[0169] Definitions i For the data packet sender, is the data packet receiving end. The data packet path selection indicator variable is a binary variable It indicates that, Indicates the timing of the rth round From i Receive the data packet, otherwise Defining binary variables It is the variable indicating the topological connection between the intelligent gateway and each terminal. Indicates i and Connect, otherwise The timing data packet transmission model, timing delay model and cumulative timing error model are considered as follows.

[0170] 1) Power line carrier timing data packet transmission model

[0171] Considering the complex electromagnetic environment and noise interference of the low-carbon smart park, the present invention realizes the transmission of power line carrier timing data packets based on orthogonal frequency division multiplexing (OFDM) technology. From i The rate at which packets are received is:

[0172]

[0173] Among them, N s is the OFDM symbol transmission rate, is the transmission power, represents the path gain. Among them, and σ0 are the path frequency response, electromagnetic interference and Gaussian white noise power respectively. represents the signal-to-interference-noise ratio gap, that is, the ability to resist signal-to-interference-noise ratio fading. e represents the target bit error rate, Q -1 (x) indicates The inverse function of .

[0174] In order to ensure the transmission quality of the power line carrier timing data packet, the present invention imposes constraints on the signal-to-interference-to-noise ratio of the path, namely:

[0175]

[0176] 2) Timing delay model

[0177] The present invention divides each round of timing into several time slots of length τ. At the beginning of each time slot, s i Start connecting Transmit power line carrier timing data packets. In the multi-hop transmission of the low-carbon smart park terminal topology, the rth round of timing The delay of receiving a data packet is determined by the sender s i The delay of receiving data packets and From i The single-hop transmission delay of the received data packet consists of two parts, expressed as

[0178]

[0179] in, Indicates i The number of time slots it takes to receive a packet, for From i The single-hop transmission delay of receiving data packets is expressed as:

[0180]

[0181] Among them, U r Indicates the size of the rth round of timing data packets.

[0182] 3) Cumulative timing error model

[0183] Define round r The cumulative timing error of the received data packet is Expressed as

[0184]

[0185] in, The timing for the rth round From i The cumulative timing error of the received data packet is

[0186]

[0187] in, Indicates the rth round of timing From i The single-hop timing error caused by receiving data packets obeys the normal distribution with a mean error of α and a standard deviation of β, that is, The rth round of timing s i Cumulative timing error of received data packets

[0188] (2) Optimization Problem Modeling

[0189] The optimization goal of the present invention is to minimize the weighted sum of the average timing delay and timing error per round by optimizing the path selection strategy of the power line carrier timing data packet under the constraint of signal-to-interference-noise ratio. The optimization problem can be modeled as

[0190]

[0191] Among them, χ is a weight parameter used to dynamically weigh the timing delay and timing error. C1 and C2 are path selection constraints, indicating the timing receiver in each round A data packet can be received from at most one sender. C3 means With s i When not connected, Unable to i Receive data packets. C4 is the signal-to-interference-noise ratio constraint.

[0192] (3) Power line carrier sensing topology prediction method for low-carbon smart park clusters

[0193] In order to achieve high-precision and low-latency timing of low-carbon smart park group terminals, it is necessary to accurately predict the topology of the park group terminals and obtain a topology adjacency matrix. Based on this, the present invention proposes a power line carrier sensing topology prediction method for low-carbon smart park groups, and inputs the historical topology information stored in the low-carbon smart park database into the neural network for pre-training. Real-time power line carrier monitoring data such as signal-to-interference-noise ratio and electromagnetic interference are input into the trained BP neural network to sense the terminal topology adjacency relationship and achieve accurate topology prediction. The schematic diagram of the power line carrier sensing topology prediction method for low-carbon smart park groups is shown in the figure. Figure 2 As shown, the steps are as follows:

[0194] Step 1: BP neural network construction and pre-training

[0195] The present invention inputs the park terminal topology adjacency matrix stored in the low-carbon smart park database, the inter-topology characteristic data such as signal-to-noise ratio and electromagnetic interference, and the historical timing delay and error data into the neural network for pre-training, and after comprehensive analysis, it simulates the relevant parameters of the park topology prediction to achieve topology prediction. The constructed neural network model is as follows:

[0196] The BP neural network includes an input layer, a hidden layer, and an output layer. The number of neurons in the input layer is X, the number of neurons in each hidden layer is Y, and the number of neurons in the output layer is Z. Among them, the hidden layer parameters are Q(A, B), Q is the BP neural network parameter matrix, A is the hidden layer weight matrix, and B is the threshold matrix.

[0197] The input signal matrix of the BP neural network input layer suitable for terminal topology prediction can be expressed as follows:

[0198]

[0199] in, Represent the input signals of the 1st, xth, and Xth input layer neurons respectively. Therefore, the activation function of the BP neural network is expressed as

[0200]

[0201] Among them, κ is the activation function parameter. The hidden layer function of the BP neural network suitable for terminal topology prediction is

[0202]

[0203] in is the weight between the input layer neurons and the hidden layer neurons, b y is the hidden layer threshold, N y is the output of the yth neuron in the hidden layer. The output layer function of the BP neural network suitable for terminal topology prediction is

[0204]

[0205] in, is the weight between the hidden layer neurons and the output layer neurons, c z is the output layer threshold, G z is the output of the zth neuron.

[0206] During the pre-training process, each training will produce errors. The loss function can reflect the gap between the model and the actual data. By back-propagating and updating the weights and biases, the next training can gradually reduce the error and improve the prediction accuracy. In order to judge the error between the estimated value and the actual value, the loss function is defined.

[0207]

[0208] Among them, M is the number of samples, k is the output of the neural network training sample, and k *is the expected output, and m is the dimension of the data. The weights are updated according to the negative gradient direction of the error. When the input sample and the expected output are given, the iteration is repeated for each input sample. When all the input sample data are iterated in sequence, after the sample training is completed, it is determined whether the indicator function meets the accuracy. If the indicator function meets the accuracy, the training is stopped, otherwise the training is repeated until the accuracy is met.

[0209] Step 2: Terminal topology adjacency matrix prediction

[0210] After obtaining the BP neural network model suitable for terminal topology prediction according to the above process, at the beginning of each round of timing process, the power line carrier monitoring data such as signal-to-noise ratio and electromagnetic interference between terminals are input into the trained BP neural network model, and the terminal topology connection relationship of this round of timing is obtained through the mapping relationship between the terminal topology adjacency relationship matrix fitted in the pre-training process and the input data. Terminal topology adjacency relationship matrix W r It can be expressed by the following formula:

[0211]

[0212] (4) A low-carbon smart park power line carrier timing method based on synchronous group game

[0213] Get the terminal topology adjacency matrix W r Finally, the present invention proposes a low-carbon smart park group power line carrier timing method based on synchronous group game to optimize the power line carrier timing data packet path selection decision. In the rth round of timing, a power line carrier timing data packet transmitter and multiple timing data packet receivers are defined as a multicast group, denoted as The utility function can be expressed as follows:

[0214]

[0215] in, for The number of data packet receivers.

[0216] Since the timing data packets of the campus terminals are transmitted through the multicast group, the optimization problem can be expressed as a synchronization group game problem. Each multicast group constitutes a synchronization group. Each terminal, as a player, tends to join the synchronization group to receive data packets to reduce its own timing delay and timing error, while not increasing the average delay and error of the receiving end of the data packet in the group. The present invention defines the synchronization group game of the rth round of timing as Game r (Φ r ,V), where Φ represents the set of terminals participating in the game, satisfying V represents the value of the synchronization group, defined as:

[0217]

[0218] In each synchronization group, whether the synchronization group allows the data packet receiver to join depends on its contribution to the value of the synchronization group. If the receiver seriously increases the average timing delay and timing error of the multicast group, it will be rejected. Benefits is defined as the contribution to the sync group value, calculated as:

[0219]

[0220] in, express Join Sync Group

[0221] In order to solve the above-mentioned synchronous group game problem, this paper adopts a synchronous group formation algorithm, the flow chart is as follows Figure 3 The specific introduction is as follows:

[0222] Step 1: Synchronization group initialization

[0223] According to the signal-to-noise ratio and topological adjacency matrix between terminals, the priority of each receiving terminal to all sending terminals is calculated, and the preference order is arranged in descending order of signal-to-noise ratio. If there is a sender that satisfies formula (2), add it to the set at the same time, In Temporarily join a sync group where s i The first priority. If the formula (2) is not satisfied between the receiving end and any sending end, then Does not participate in this round of timing.

[0224] Step 2: Create a Sync Group

[0225] When all synchronization groups in this time slot are temporarily established, the receiving end, e.g. You can choose to join other synchronization groups based on the benefits calculated by formula (16). but Select Leave current sync group Join the sync group The receiving end in the process can continuously choose to change the synchronization group according to the benefits until no receiving end chooses to change the synchronization group.

[0226] Step 3: Final synchronization group establishment and power line carrier timing data packet transmission

[0227] Repeat step 2 until all receiving ends do not change their selection. The present invention calls the time when the synchronization group is finally formed a decision point. After this, the transmitting end in each synchronization group starts to transmit the power line carrier timing data packet in the same synchronization group. When each transmitting end in all synchronization groups completes the timing, the synchronization group of this time slot is disbanded.

[0228] Step 4: Next time slot transmission

[0229] At the end of each time slot, the gateway checks whether all synchronization groups have completed transmission. Then, the receiving end that receives the power line carrier timing data packet in the current time slot will serve as the sending end for the next time slot transmission. The algorithm goes to step 1.

[0230] When each terminal completes timing synchronization, this round of timing synchronization ends.

[0231] Embodiment 2:

[0232] In order to better implement the above method, this embodiment provides a low-carbon park cluster topology prediction and timing system based on power line carrier, and the system includes:

[0233] The system is applied to the control layer, and the system includes:

[0234] The first building module is used to build a power line carrier timing system model by analyzing the actual distribution of smart gateways and monitoring terminals in the smart park group and the precise timing information data packets;

[0235] The second building module is used to construct an optimization problem by optimizing the path selection strategy of the power line carrier timing data packet based on the power line carrier timing system model under the signal-to-interference-noise ratio constraint;

[0236] The topology prediction module is used to obtain the historical topology information and real-time monitoring data of the power line carrier stored in the low-carbon smart park database, predict the topology of the monitoring terminal, and obtain the topological adjacency relationship of the monitoring terminal;

[0237] The decision optimization module is used to optimize the decision of the timing information data packet path based on the signal-to-interference-noise ratio between monitoring terminals in the real-time monitoring data of the power line carrier, the topological adjacency relationship of the monitoring terminals and the constructed optimization problem, and obtain the optimal timing decision result by using the synchronization group game algorithm. The decision result realizes the time synchronization of the park group monitoring terminals and electrical equipment.

[0238] Further, the first building module includes: an analysis unit and a first building unit;

[0239] The analysis unit is used to determine that there is an intelligent gateway and I monitoring terminals in the park, and determine the set and express it as S = {s0, s1, ..., s i ,...,sI}, where s0 represents the smart gateway, s1, s i and I Respectively represent the 1st, i, and Ith monitoring terminals; if the connection topology of the monitoring terminals in the park is unknown, a total of R rounds of timing processes are considered, and the set is determined and expressed as R = {1, ..., r, ..., R}, where 1, r, and R represent the 1st, r, and R rounds of timing, respectively;

[0240] Determine i For the data packet sender, The packet receiving end; the packet path selection indicator variable is a binary variable It indicates that, Indicates the timing of the rth round From i Receive the data packet, otherwise Identify binary variables is the topological connection indicator variable between the intelligent gateway and each monitoring terminal. Indicates i and Connect, otherwise

[0241] Further, the first construction unit is used to construct the power line carrier timing data packet transmission model, specifically:

[0242] Considering the complex electromagnetic environment and noise interference of the low-carbon smart park, the transmission of power line carrier timing data packets is realized based on orthogonal frequency division multiplexing technology. From i The rate at which packets are received is:

[0243]

[0244] Among them, N s is the OFDM symbol transmission rate, is the transmission power, represents the path gain; where and σ0 are the path frequency response, electromagnetic interference and Gaussian white noise power respectively; represents the signal-to-interference-noise ratio gap, that is, the ability to resist signal-to-interference-noise ratio fading; where P e represents the target bit error rate, Q -1 (x) indicates The inverse function of .

[0245] Used to construct the timing delay model; specifically:

[0246] Each round of timing is divided into several time slots of length τ. At the beginning of each time slot, s i Start connecting Transmit power line carrier timing data packets; in the multi-hop transmission of the low-carbon smart park monitoring terminal topology, the rth round of timing The delay of receiving a data packet is determined by the sender s i The delay of receiving data packets and From i The single-hop transmission delay of receiving a data packet consists of two parts, expressed as:

[0247]

[0248] in, Indicates i The number of time slots it takes to receive a packet, for From i The single-hop transmission delay of receiving data packets is expressed as:

[0249]

[0250] Among them, U r Indicates the size of the rth round of timing data packets.

[0251] Used to construct the cumulative timing error model, specifically:

[0252] Determine the rth round Cumulative timing error of received data packets for:

[0253]

[0254] in, The timing for the rth round From i The cumulative timing error of the received data packet is:

[0255]

[0256] in, Indicates the rth round of timing From i The single-hop timing error caused by receiving data packets obeys the normal distribution with a mean error of α and a standard deviation of β, that is, The rth round of timing s i Cumulative timing error of received data packets.

[0257] Furthermore, the second construction module includes: constructing an optimization problem unit;

[0258] The optimization problem unit is used to minimize the weighted sum of the average timing delay and timing error per round under the constraint of signal-to-interference-noise ratio by optimizing the path selection strategy of the power line carrier timing data packet. The optimization problem is specifically constructed as follows:

[0259]

[0260] Among them, χ is a weight parameter used to dynamically weigh the timing delay and timing error; C1 and C2 are path selection constraints, indicating the timing of each round of the receiving end. Can receive data packets from at most one sender; C3 means With s i When not connected, Unable to i Receive data packets; To ensure the transmission quality of the power line carrier timing data packets, a constraint is imposed on the signal-to-interference-to-noise ratio of the path, namely C4:

[0261]

[0262] Furthermore, the topology prediction module includes:

[0263] A monitoring unit, configured to select a timing path for a precise timing information data packet based on a monitoring terminal topology adjacency relationship;

[0264] A pre-training unit is used to input the historical topology information into a preset BP neural network for pre-training to obtain a trained BP neural network model;

[0265] The inputting of the historical topology information into a preset BP neural network for pre-training specifically includes:

[0266] The topological adjacency matrix of the park monitoring terminals, the characteristic data between topologies, the historical power line carrier timing delay and the historical power line carrier timing error stored in the low-carbon smart park database are input into the preset BP neural network for pre-training;

[0267] The preset BP neural network is constructed specifically as follows:

[0268] The BP neural network consists of an input layer, a hidden layer, and an output layer. The number of neurons in the input layer is X, the number of neurons in each hidden layer is Y, and the number of neurons in the output layer is Z.

[0269] Among them, the hidden layer parameters are Q(A,B), Q is the BP neural network parameter matrix, A is the hidden layer weight matrix, and B is the threshold matrix;

[0270] The input signal matrix of the BP neural network input layer suitable for monitoring terminal topology prediction is expressed as:

[0271]

[0272] in, Represent the input signals of the 1st, xth, and Xth input layer neurons respectively;

[0273] The activation function of the BP neural network is expressed as:

[0274]

[0275] Among them, κ is the activation function parameter; the hidden layer function of the BP neural network suitable for monitoring terminal topology prediction is:

[0276]

[0277] in, is the weight between the input layer neurons and the hidden layer neurons, b y is the hidden layer threshold, N y is the output of the yth neuron in the hidden layer;

[0278] The output layer function of the BP neural network suitable for monitoring terminal topology prediction is:

[0279]

[0280] in, is the weight between the hidden layer neurons and the output layer neurons, c z is the output layer threshold, G z is the output of the zth neuron;

[0281] The loss function of the BP neural network is expressed as:

[0282]

[0283] Among them, M is the number of samples, k is the output of the neural network training sample, and k * is the expected output, m is the dimension of the data; the weights are updated according to the negative gradient direction of the error; when the input signal matrix and the expected output are given, each input signal matrix is ​​iterated repeatedly, and when all input signal matrix data are iterated in turn, after the sample training is completed, it is determined whether the indicator function meets the accuracy; if the indicator function meets the accuracy, the training is stopped, otherwise the training is repeated until the accuracy is met, and a trained BP neural network model suitable for monitoring terminal topology prediction is obtained;

[0284] A solving unit is used to input the real-time monitoring data of the power line carrier into the trained BP neural network model based on the constructed power line carrier timing system model to obtain the topological adjacency relationship of the monitoring terminal;

[0285] The real-time monitoring data of the power line carrier is input into the trained BP neural network model to obtain the topological adjacency relationship of the monitoring terminal, which specifically includes:

[0286] At the beginning of each round of timing process, the real-time monitoring data of the power line carrier between the monitoring terminals is input into the trained BP neural network model to obtain the topological connection relationship of the monitoring terminals in this round of timing; the monitoring terminal topological adjacency relationship matrix W r It can be expressed by the following formula:

[0287]

[0288] Furthermore, the decision optimization module includes:

[0289] Synchronization group initialization unit: used to calculate the priority of each receiving end to the sender of all received timing information data packets according to the signal interference and noise ratio and topological adjacency between the monitoring terminals. The preference order is arranged in descending order of signal interference and noise ratio. If there is a satisfying The sender of the at the same time, In Temporarily join a sync group where s i First in priority; if the receiving end Does not satisfy any sender The receiving end Does not participate in this round of timing;

[0290] Establishing synchronization group unit: used when all synchronization groups of this time slot are temporarily established, the receiving end, for example According to the formula The calculated benefit is the optimization goal to choose to join other synchronization groups; if but Select Leave current sync group Join the sync group The receiving end in can continuously choose to change the synchronization group according to the benefits until no receiving end chooses to change the synchronization group;

[0291] Transmission establishment unit, used for final synchronization group establishment and power line carrier timing data packet transmission: specifically:

[0292] Repeat the processing method of establishing the synchronization group unit until all receiving ends do not change their selection; the time when the synchronization group is finally formed is called the decision point; after this, the transmitting end in each synchronization group starts to transmit the power line carrier timing data packet in the same synchronization group; when each transmitting end in all synchronization groups completes the timing, the synchronization group of this time slot is disbanded;

[0293] The detection unit is used to check whether all synchronization groups have completed transmission at the end of each time slot; then, the receiving end that receives the power line carrier timing data packet in the current time slot will serve as the sending end for the next time slot transmission; the algorithm is transferred to the synchronization group initialization unit; when each monitoring terminal completes the timing, this round of timing ends.

[0294] Embodiment 3:

[0295] A person of ordinary skill in the art will appreciate that all or part of the steps in the various methods of the above embodiments may be completed by instructions, or by controlling related hardware through instructions. The instructions may be stored in a computer-readable storage medium and loaded and executed by a processor.

[0296] To this end, an embodiment of the present invention provides a storage medium storing a plurality of instructions, which can be loaded by a processor to execute the steps of any one of the low-carbon campus group topology prediction and timing methods based on power line carrier provided in an embodiment of the present invention.

[0297] For example, the instruction can execute the following steps:

[0298] By analyzing the actual distribution of smart gateways and monitoring terminals in the smart park cluster and the precise timing information data packets, a power line carrier timing system model is constructed;

[0299] Based on the power line carrier timing system model, under the constraint of signal-to-interference-noise ratio, the optimization problem is constructed by optimizing the path selection strategy of the power line carrier timing data packet.

[0300] Obtain the historical topology information and real-time monitoring data of power line carrier stored in the low-carbon smart park database, predict the topology of monitoring terminals, and obtain the topological adjacency of monitoring terminals;

[0301] Based on the signal-to-interference-to-noise ratio between monitoring terminals in the real-time monitoring data of the power line carrier, the topological adjacency relationship of the monitoring terminals and the optimization problem constructed, the decision of the timing information data packet path is optimized by using the synchronization group game algorithm to obtain the optimal timing decision result, and the decision result realizes the time synchronization of the park group monitoring terminals and electrical equipment. The preferred embodiments of the present invention are described in detail above, but the present invention is not limited to the above embodiments, and various changes can be made within the knowledge of ordinary technicians in this field without departing from the purpose of the present invention.

[0302] Those skilled in the art will appreciate that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can adopt the form of complete hardware embodiments, complete software embodiments, or embodiments in combination with software and hardware. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code. The scheme in the embodiments of the present application can be implemented in various computer languages, for example, object-oriented programming language Java and literal scripting language JavaScript, etc.

[0303] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0304] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 These computer program instructions can also be loaded onto a computer or other programmable data processing device, so that a series of operation steps are executed on the computer or other programmable device to produce a computer-implemented process, so that the instructions executed on the computer or other programmable device provide for implementing the process in the process. Figure 1 A process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0305] Although the preferred embodiments of the present application have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications falling within the scope of the present application.

[0306] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalents, the present application is also intended to include these modifications and variations.

Claims

1. A low-carbon park cluster topology prediction and timing method based on power line carrier, characterized in that: The method comprises: By analyzing the actual distribution of smart gateways and monitoring terminals in the smart park cluster and the precise timing information data packets, a power line carrier timing system model is constructed; The analysis of the actual distribution of smart gateways and monitoring terminals in the smart park group and the precise timing information data packets is specifically as follows: Analyze the actual distribution of smart gateways and monitoring terminals in the smart park cluster, determine the set and express it as S = {s0, s1, ..., s i ,...,s I }, where s0 represents the smart gateway, s1, s i and I Respectively represent the 1st, i, and Ith monitoring terminals; if the connection topology of the monitoring terminals in the park is unknown, a total of R rounds of timing processes are considered, and the set is determined and expressed as R = {1, ..., r, ..., R}, where 1, r, and R represent the 1st, r, and R rounds of timing, respectively; Analyze the precise timing information data packet to determine s i For the data packet sender, is the data packet receiving end; the data packet path selection indicator variable is converted into a binary variable It indicates that, Indicates the timing of the rth round From i Receive the data packet, otherwise Identify binary variables is the topological connection indicator variable between the intelligent gateway and each monitoring terminal. Represented as s i and Connect, otherwise Based on the power line carrier timing system model, under the constraint of signal-to-interference-noise ratio, the optimization problem is constructed by optimizing the path selection strategy of the power line carrier timing data packet. Obtain the historical topology information and real-time monitoring data of power line carrier stored in the low-carbon smart park database, predict the topology of monitoring terminals, and obtain the topological adjacency of monitoring terminals; Based on the signal-to-interference-to-noise ratio between monitoring terminals in the real-time monitoring data of the power line carrier, the topological adjacency relationship of the monitoring terminals and the optimization problem constructed, the synchronous group game algorithm is used to optimize the decision of the timing information data packet path to obtain the optimal timing decision result, which realizes the time synchronization of the park group monitoring terminals and electrical equipment; The power line carrier timing system model includes: a power line carrier timing data packet transmission model, a timing delay model and a cumulative timing error model; The power line carrier timing data packet transmission model is constructed as follows: Considering the complex electromagnetic environment and noise interference of the low-carbon smart park, the transmission of power line carrier timing data packets is realized based on orthogonal frequency division multiplexing technology. From i The rate at which packets are received is: in, is a binary variable, N s is the OFDM symbol transmission rate, is the transmission power, represents the path gain; where and σ0 are the path frequency response, electromagnetic interference and Gaussian white noise power respectively; represents the signal-to-interference-noise ratio gap, that is, the ability to resist signal-to-interference-noise ratio fading; where P e represents the target bit error rate, Q -1 (x) indicates The inverse function of Construct the timing delay model; specifically: Each round of timing is divided into several time slots of length τ. At the beginning of each time slot, s i Start connecting Transmit power line carrier timing data packets; in the multi-hop transmission of the low-carbon smart park monitoring terminal topology, the rth round of timing The delay of receiving a data packet is determined by the sender s i The delay of receiving data packets and From i The single-hop transmission delay of receiving a data packet consists of two parts, expressed as: in, Indicates that in the rth round of timing, the node Receive from node s i The total delay of the data packet, Indicates i The number of time slots it takes to receive a packet, for From i The single-hop transmission delay of receiving data packets is expressed as: Among them, U r Indicates the size of the rth round of timing data packets, Indicates the transmission rate of timing data packets; Construct the cumulative timing error model, specifically: Determine the rth round Cumulative timing error of received data packets for: in, is a binary variable, The timing for the rth round From i The cumulative timing error of the received data packet is: in, Indicates the rth round of timing From i The single-hop timing error caused by receiving data packets obeys the normal distribution with a mean error of α and a standard deviation of β, that is, The rth round of timing s i Cumulative timing error of received data packets.

2. The method for predicting and timing low-carbon park cluster topology based on power line carrier according to claim 1 is characterized in that: The optimization goal is to minimize the weighted sum of the average timing delay and timing error per round by optimizing the path selection strategy of the power line carrier timing data packet under the constraint of signal-to-interference-noise ratio. The optimization problem is specifically constructed as follows: Among them, χ is a weight parameter used to dynamically weigh the timing delay and timing error; C1 and C2 are path selection constraints, indicating the timing of each round of the receiving end. Receive data packets from at most one sender; C3 means With s i When not connected, Unable to i Receive data packets; To ensure the transmission quality of the power line carrier timing data packets, a constraint is imposed on the signal-to-interference-to-noise ratio of the path, namely C4: in, Indicates i and The signal-to-interference-noise ratio between min In order to ensure the signal-to-noise ratio threshold of transmission quality in the power line carrier channel under the influence of complex electromagnetic environment and noise interference in the park, C4 indicates that only when the signal-to-noise ratio of the path is greater than the signal-to-noise ratio threshold ζ min Only when the path is in a topological connection state can it be determined that the path is in a topological connection state.

3. The method for predicting and timing low-carbon park cluster topology based on power line carrier according to claim 1 is characterized in that: The monitoring terminal topology prediction specifically includes: Select a timing path for precise timing information packets based on the topological adjacency of the monitoring terminals; Input the historical topology information into the preset BP neural network for pre-training to obtain a trained BP neural network model; Based on the constructed power line carrier timing system model, the real-time monitoring data of the power line carrier is input into the trained BP neural network model to obtain the topological adjacency relationship of the monitoring terminals.

4. The method for predicting and timing low-carbon park cluster topology based on power line carrier according to claim 3 is characterized in that: The inputting of the historical topology information into a preset BP neural network for pre-training specifically includes: The topological adjacency matrix of the park monitoring terminals, the characteristic data between topologies, the historical power line carrier timing delay and the historical power line carrier timing error stored in the low-carbon smart park database are input into the preset BP neural network for pre-training; The preset BP neural network is constructed specifically as follows: The BP neural network consists of an input layer, a hidden layer, and an output layer. The number of neurons in the input layer is X, the number of neurons in each hidden layer is Y, and the number of neurons in the output layer is Z. Among them, the hidden layer parameters are Q(A,B), Q is the BP neural network parameter matrix, A is the hidden layer weight matrix, and B is the threshold matrix; The input signal matrix of the BP neural network input layer suitable for monitoring terminal topology prediction is expressed as: in, Represent the input signals of the 1st, xth, and Xth input layer neurons respectively; The activation function of the BP neural network is expressed as: Among them, κ is the activation function parameter; the hidden layer function of the BP neural network suitable for monitoring terminal topology prediction is: in, is the weight between the input layer neurons and the hidden layer neurons, b y is the hidden layer threshold, N y is the output of the yth neuron in the hidden layer; The output layer function of the BP neural network suitable for monitoring terminal topology prediction is: in, is the weight between the hidden layer neurons and the output layer neurons, c z is the output layer threshold, G z is the output of the zth neuron; The loss function of the BP neural network is expressed as: Among them, M is the number of samples, k is the output of the neural network training sample, and k * is the expected output, m is the dimension of the data; the weights are updated according to the negative gradient direction of the error; when the input signal matrix and the expected output are given, each input signal matrix is ​​iterated repeatedly, and when all the input signal matrix data are iterated repeatedly in turn, after the sample training is completed, it is determined whether the indicator function meets the accuracy; if the indicator function meets the accuracy, the training is stopped, otherwise, the training is repeated until the accuracy is met, and a trained BP neural network model suitable for monitoring terminal topology prediction is obtained.

5. The method for predicting and timing low-carbon park cluster topology based on power line carrier according to claim 4 is characterized in that: The real-time monitoring data of the power line carrier is input into the trained BP neural network model to obtain the topological adjacency relationship of the monitoring terminal, which specifically includes: At the beginning of each round of timing process, the real-time monitoring data of the power line carrier between the monitoring terminals is input into the trained BP neural network model to obtain the topological connection relationship of the monitoring terminals in this round of timing; the monitoring terminal topological adjacency relationship matrix W r It can be expressed by the following formula:

6. The method for predicting and timing low-carbon park cluster topology based on power line carrier according to claim 1 is characterized in that: Before optimizing the decision of the timing information data packet path by using the synchronization group game algorithm, the method further includes: First, in the rth round of timing, a power line carrier timing data packet transmitter and multiple timing data packet receivers are defined as a multicast group, denoted as The utility function can be expressed as follows: in, for The number of data packet receivers in The synchronization group game of the rth round of timing is defined as Game r (Φ r ,V), where Φ represents the set of monitoring terminals participating in the game, satisfying V represents the value of the synchronization group, defined as: In each synchronization group, whether the synchronization group allows the data packet receiver to join depends on its contribution to the value of the synchronization group. If the receiver seriously increases the average timing delay and timing error of the multicast group, it will be rejected; therefore, the receiver in the timing synchronization group of the rth round Benefits is defined as the contribution to the sync group value, calculated as: in, express Join Sync Group 7. The method for predicting and timing low-carbon park cluster topology based on power line carrier according to claim 6 is characterized in that: The synchronization group game algorithm is used to optimize the decision of the timing information data packet path, specifically: Step 1: Synchronization group initialization: According to the signal-to-interference-noise ratio and topological adjacency between the monitoring terminals, the priority of each receiving terminal to the senders of all received timing information packets is calculated, and the preference order is arranged in descending order of signal-to-interference-noise ratio; for the receiving terminal If there is a satisfying The sender of the at the same time, In Temporarily join a sync group where s i First in priority; if the receiving end Does not satisfy any sender The receiving end Does not participate in this round of timing; Step 2: Create a sync group: When all synchronization groups in this time slot are temporarily established, the receiving end, for example According to the formula The calculated benefit is the optimization goal to choose to join other synchronization groups; if but Select Leave current sync group Join the sync group The receiving end in can continuously choose to change the synchronization group according to the benefits until no receiving end chooses to change the synchronization group; Step 3: Final synchronization group establishment and power line carrier timing data packet transmission: Repeat step 2 until all receiving ends do not change their selections; the time when the synchronization group is finally formed is called the decision point; after this, the transmitter in each synchronization group starts to transmit the power line carrier timing data packet in the same synchronization group; when each transmitter in all synchronization groups completes the timing, the synchronization group of this time slot is disbanded; Step 4: Next time slot transmission: At the end of each time slot, check whether all synchronization groups have completed transmission; then, the receiving end that receives the power line carrier timing data packet in the current time slot will serve as the sending end for the next time slot transmission; the algorithm goes to step 1; when each monitoring terminal completes timing, this round of timing ends.

8. A low-carbon park cluster topology prediction and timing system based on power line carrier, characterized in that: The system is applied to the control layer, and the system includes: The first building module is used to build a power line carrier timing system model by analyzing the actual distribution of smart gateways and monitoring terminals in the smart park group and the precise timing information data packets; Specifically, the actual distribution of smart gateways and monitoring terminals in the smart park cluster is analyzed, and the set is determined and expressed as S = {s0, s1, ..., s i ,...,s I }, where s0 represents the smart gateway, s1, si and s I Respectively represent the 1st, i, and Ith monitoring terminals; if the connection topology of the monitoring terminals in the park is unknown, a total of R rounds of timing processes are considered, and the set is determined and expressed as R = {1, ..., r, ..., R}, where 1, r, and R represent the 1st, r, and R rounds of timing, respectively; Analyze the precise timing information data packet to determine s i For the data packet sender, is the data packet receiving end; the data packet path selection indicator variable is converted into a binary variable It indicates that, Indicates the timing of the rth round From i Receive the data packet, otherwise Identify binary variables is the topological connection indicator variable between the intelligent gateway and each monitoring terminal. Represented as s i and Connect, otherwise The second building module is used to construct an optimization problem by optimizing the path selection strategy of the power line carrier timing data packet based on the power line carrier timing system model under the signal-to-interference-noise ratio constraint; The power line carrier timing system model includes: a power line carrier timing data packet transmission model, a timing delay model and a cumulative timing error model; The power line carrier timing data packet transmission model is constructed as follows: Considering the complex electromagnetic environment and noise interference of the low-carbon smart park, the transmission of power line carrier timing data packets is realized based on orthogonal frequency division multiplexing technology. From i The rate at which packets are received is: Among them, N s is the OFDM symbol transmission rate, is the transmission power, represents the path gain; where and σ0 are the path frequency response, electromagnetic interference and Gaussian white noise power respectively; represents the signal-to-interference-noise ratio gap, that is, the ability to resist signal-to-interference-noise ratio fading; where P e represents the target bit error rate, Q -1 (x) indicates The inverse function of Construct the timing delay model; specifically: Each round of timing is divided into several time slots of length τ. At the beginning of each time slot, s i Start connecting Transmit power line carrier timing data packets; in the multi-hop transmission of the low-carbon smart park monitoring terminal topology, the rth round of timing The delay of receiving a data packet is determined by the sender s i The delay of receiving data packets and From i The single-hop transmission delay of receiving a data packet consists of two parts, expressed as: in, Indicates that in the rth round of timing, the node Receive from node s i The total delay of the data packet, Indicates i The number of time slots it takes to receive a packet, for From i The single-hop transmission delay of receiving data packets is expressed as: Among them, among them, Indicates that in the rth round of timing, the node Receive from node s i The total delay of the data packet, U r Indicates the size of the rth round of timing data packet; Construct the cumulative timing error model, specifically: Determine the rth round Cumulative timing error of received data packets for: in, The timing for the rth round From i The cumulative timing error of the received data packet is: in, Indicates the rth round of timing From i The single-hop timing error caused by receiving data packets obeys the normal distribution with a mean error of α and a standard deviation of β, that is, The rth round of timing s i The accumulated timing error of the received data packets; The topology prediction module is used to obtain the historical topology information and real-time monitoring data of the power line carrier stored in the low-carbon smart park database, predict the topology of the monitoring terminal, and obtain the topological adjacency relationship of the monitoring terminal; The decision optimization module is used to optimize the decision of the timing information data packet path based on the signal-to-interference-noise ratio between monitoring terminals in the real-time monitoring data of the power line carrier, the topological adjacency relationship of the monitoring terminals and the constructed optimization problem, and obtain the optimal timing decision result by using the synchronization group game algorithm. The decision result realizes the time synchronization of the park group monitoring terminals and electrical equipment.