Power transmission line dynamic capacity increase intelligent regulation and control method and system

Abstract

The invention discloses an intelligent regulation and control method and system for dynamic capacity increase of a power transmission line, and belongs to the technical field of power system regulation and control and data fusion. The method specifically comprises the following steps: collecting and preprocessing measurement data of nodes on a power transmission line, fusing the preprocessed measurement data by using Kalman filtering, a Bayesian network and a multi-sensor fusion algorithm, constructing a dynamic heat balance model and a dynamic load evaluation model, and calculating a real-time operation state and a safety margin of a power grid; inputting a data combination obtained after multivariate fusion with measurement data of nodes on the power transmission line at the past L moment into an intelligent decision-making algorithm, and constructing an intelligent decision-making prediction model; and after multivariate fusion is carried out on measurement data of nodes on a real-time power transmission line, the measurement data is combined with a real-time operation state and a safety margin of a power grid, an intelligent decision prediction model is input, and a global collaborative dynamic capacity increasing control strategy is generated according to a prediction result. According to the invention, the power grid resource utilization rate and the new energy consumption capability are improved, and safe, stable and efficient operation of the power grid under complex operation conditions is guaranteed.

Description

Technical Field

[0001] This invention belongs to the field of power system regulation and data fusion technology, and more specifically, relates to an intelligent management system for dynamic capacity expansion of transmission lines. By collecting multi-source data, including renewable energy generation, environmental data, load data, and equipment status, and utilizing data fusion and intelligent decision-making algorithms, this system enables real-time management and regulation of dynamic grid capacity expansion. This aims to improve renewable energy absorption capacity and grid resource utilization efficiency, while ensuring safe and stable system operation. Background Art

[0002] Dynamic capacity expansion management technology for new energy power grids enables comprehensive assessment of the grid's real-time operating status through online collection of multiple data sources, including renewable energy generation data, environmental parameters, user loads, and key equipment status. Based on multi-dimensional data collected on-site and various safety operating criteria, this technology calculates and determines the grid's current steady-state operating capacity limit in real time.

[0003] Existing technologies often focus on single data sources or single-point monitoring, such as focusing solely on ambient temperature or load data, and determining the maximum capacity of the power grid through traditional thermal equations and load flow calculations. However, with the large-scale integration of renewable energy sources such as wind and solar power, their power generation has become significantly intermittent and volatile, making single-data monitoring methods unable to fully reflect the actual operating status of the power grid.

[0004] Existing technologies process each data source independently without considering the correlation between the data. This makes it impossible to determine the actual operating status of the power grid in complex scenarios, which easily leads to a high misjudgment rate of the actual operating status of the power grid. Summary of the Invention

[0005] To address the deficiencies in the prior art, the present invention provides an intelligent management system for dynamic capacity expansion of transmission lines. By integrating multi-source real-time data from new energy monitoring, environmental monitoring, load detection, and equipment status collection, a unified data model is constructed. An intelligent decision-making algorithm is used to conduct real-time evaluation of the grid operation status, generate a global dynamic capacity expansion control strategy, and thereby achieve automatic regulation and optimization of the dynamic capacity of the grid.

[0006] The present invention adopts the following technical solutions.

[0007] A first aspect of the present invention provides a method for intelligently controlling dynamic capacity expansion of a transmission line, specifically comprising:

[0008] Collect measurement data on new energy, environment, load and equipment status at nodes on the transmission line;

[0009] Preprocessing the measurement data collected from nodes on the transmission line to obtain preprocessed measurement data;

[0010] Using Kalman filtering, Bayesian network and multi-sensor fusion algorithm, the pre-processed measurement data is fused into multivariate fusion state data;

[0011] Based on the multivariate fusion state data, a dynamic heat balance model and a dynamic load assessment model are constructed, and the real-time operation status and safety margin of the power grid are calculated based on the dynamic capacity assessment model;

[0012] The measurement data of the nodes on the transmission line at the past L time points are multi-element fused to obtain historical fused data. The historical fused data, the real-time operation status of the power grid, and the safety margin are input into the intelligent decision-making algorithm. The intelligent decision-making prediction model is constructed using the historical real load and renewable energy power generation as labels.

[0013] After the multi-dimensional fusion of the measurement data of the nodes on the real-time transmission line, it is combined with the current real-time operating status and safety margin of the current power grid, and input into the constructed intelligent decision-making prediction model to predict the power grid load and renewable energy power generation. Based on the power grid load and renewable energy power generation, a global coordinated dynamic capacity expansion control strategy is generated to realize intelligent regulation of dynamic capacity expansion of transmission lines.

[0014] Preferably, the method of using Kalman filtering, Bayesian network and multi-sensor fusion algorithm to fuse the pre-processed measurement data into multivariate fusion state data specifically includes:

[0015] Perform Kalman filter optimal state estimation on pre-processed measurement data;

[0016] Based on the optimal state estimation, a Bayesian network is constructed and probabilistic reasoning is performed to obtain the posterior probability distribution of the Bayesian network output;

[0017] According to the posterior probability distribution, the maximum a posteriori estimate is calculated to generate multivariate fusion state data.

[0018] Preferably, the step of constructing a dynamic heat balance model and a dynamic load assessment model based on the multivariate fusion state data, and calculating the real-time operation state and safety margin of the power grid based on the dynamic capacity assessment model, specifically includes:

[0019] Based on the multivariate fusion state data, a dynamic heat balance model is constructed to obtain the solar heat absorption term, convective heat dissipation term, radiative heat dissipation term and evaporative cooling term;

[0020] Based on the obtained solar heat absorption term, convection heat dissipation term, radiation heat dissipation term and evaporative cooling term, a load assessment model is constructed to solve the current maximum safe current carrying capacity. The current maximum safe current carrying capacity is used as the upper limit of the dynamic load current to construct a dynamic load assessment model.

[0021] The real-time operation status of the power grid is calculated based on the dynamic load assessment model, and the current real-time operation status and safety margin of the power grid are solved.

[0022] Preferably, the load assessment model is constructed based on the obtained solar heat absorption term, convection heat dissipation term, radiation heat dissipation term and evaporative cooling term, the current maximum safe current carrying capacity is solved, and the current maximum safe current carrying capacity is used as the upper limit of the dynamic load current to construct a dynamic load assessment model, which specifically includes:

[0023] The solar heat absorption term, convection heat dissipation term, radiation heat dissipation term, and evaporative cooling term are summed, and the square root of the summation result and the ratio of the resistance at the maximum allowable temperature are taken to construct a load assessment model. The load assessment model is solved to obtain the current maximum safe current carrying capacity.

[0024] Subtract the theoretical maximum load of the line from the real-time load data, and multiply the subtraction result by the load margin coefficient to obtain the load multiplication result;

[0025] Multiply the real-time renewable energy power generation data by the renewable energy power generation contribution weight to obtain the renewable energy power generation multiplication result;

[0026] Multiply the real-time device status data by the device status coefficient to obtain the device status multiplication result;

[0027] The dynamic load current is obtained by adding the load multiplication result and the new energy power generation multiplication result, and then subtracting the equipment status multiplication result. The upper limit of the dynamic load current is set to the current maximum safe current carrying capacity to construct a dynamic load assessment model.

[0028] Preferably, the calculation of the real-time operation state of the power grid based on the dynamic load assessment model and the solution of the current real-time operation state of the power grid and the safety margin specifically include:

[0029] The dynamic load current obtained by solving the dynamic load assessment model, the conductor temperature in the multivariate fusion state data, and the load current in the multivariate fusion state data are combined to obtain the real-time operation status of the power grid;

[0030] The dynamic load current obtained by solving the dynamic load assessment model is subtracted from the load current in the multivariate fusion state data to obtain the safety margin.

[0031] Preferably, the measurement data of the nodes on the transmission line at the past L time points are multivariately integrated to obtain historical integrated data, the historical integrated data, the real-time operation status of the power grid and the safety margin are input into the intelligent decision-making algorithm, and the historical real load and renewable energy power generation are used as labels to construct an intelligent decision-making prediction model, which specifically includes:

[0032] The measurement data of the nodes on the transmission line at the past L time points are multivariately integrated to obtain historical fusion data, and the historical fusion data, the real-time operation status of the power grid and the safety margin are combined to obtain the decision feature vector;

[0033] The decision feature vector is input into the intelligent decision algorithm in the intelligent decision prediction model to output the load forecast and renewable energy power generation forecast results at the next moment. The intelligent decision algorithm includes convolutional neural networks, long-short-term neural networks, and support vector machines.

[0034] The current real load and renewable energy power generation are used as labels for the intelligent decision-making prediction model. The load forecast error and renewable energy power generation error are solved through the labels and prediction results. When the load forecast error and renewable energy power generation error are both greater than the set threshold, the stochastic gradient descent method is used to update the parameters of the intelligent decision-making prediction model. The parameters are adaptively updated until the load forecast error and renewable energy power generation error are both less than the set threshold, and the training of the intelligent decision-making prediction model is completed.

[0035] Preferably, the intelligent decision algorithm inputting the decision feature vector into the intelligent decision prediction model to output the load forecast and new energy power generation forecast results at the next moment specifically includes:

[0036] The decision feature vector is input into the convolutional neural network in the intelligent decision prediction model, and the spatiotemporal feature is extracted through the one-dimensional convolution layer to obtain the convolution output;

[0037] The convolution output is passed through the long-term and short-term neural network in the intelligent decision-making prediction model to extract the hidden state from the spatiotemporal features;

[0038] The hidden state is spliced ​​with the environmental data of the transmission line nodes at this moment, and then input into the support vector machine in the intelligent decision-making prediction model to perform residual correction on the splicing result, and output the load forecast and renewable energy power generation forecast results at the next moment.

[0039] Preferably, the current real load and renewable energy power generation are used as labels of the intelligent decision-making prediction model, and the load forecast error and renewable energy power generation error are solved by using the labels and prediction results. When the load forecast error and renewable energy power generation error are both greater than a set threshold, the parameters of the intelligent decision-making prediction model are updated using the stochastic gradient descent method. The parameters are adaptively updated until the load forecast error and renewable energy power generation error are both less than the set threshold, and the training of the intelligent decision-making prediction model is completed. Specifically, the training includes:

[0040] The current real load and renewable energy power generation are used as labels for the intelligent decision-making prediction model. The load prediction error is solved by taking the difference between the current real load and the load prediction result, and the renewable energy power generation error is solved by taking the difference between the current real renewable energy power generation and the renewable energy power generation prediction result.

[0041] Compare the load forecast error with the load error threshold to obtain a load comparison result, and compare the new energy power generation error with the new energy power generation threshold to obtain a new energy comparison result;

[0042] When the load comparison result and the new energy comparison result are respectively greater than the set load error threshold and new energy generation threshold, the parameters of the intelligent decision-making prediction model are updated using the stochastic gradient descent method;

[0043] The iteration is continued until the load forecast error and the renewable energy power generation error are both smaller than the set load error threshold and renewable energy power generation threshold respectively. The iteration is terminated and the training of the intelligent decision-making prediction model is completed.

[0044] Preferably, the method of updating the parameters of the intelligent decision-making prediction model using the stochastic gradient descent method specifically includes:

[0045] Calculate the mean square error loss based on the load forecast error and the renewable energy generation error;

[0046] The stochastic gradient descent method is used to calculate the gradient of the mean square error loss with respect to the parameters in the intelligent decision-making prediction model. The parameters are updated accordingly based on the gradient until the load forecast error and the renewable energy power generation error are both less than the set threshold. The iteration is terminated and the training of the intelligent decision-making prediction model is completed.

[0047] A second aspect of the present invention provides a transmission line dynamic capacity expansion intelligent control system, which runs the transmission line dynamic capacity expansion intelligent control method described in the first aspect, including:

[0048] Data acquisition module, used to collect measurement data on new energy, environment, load and equipment status of nodes on the transmission line;

[0049] A preprocessing module is used to preprocess the measurement data collected from the nodes on the transmission line to obtain preprocessed measurement data;

[0050] Data fusion module, used to fuse pre-processed measurement data into multivariate fusion state data using Kalman filtering, Bayesian network and multi-sensor fusion algorithm;

[0051] The real-time state solving module of the power grid is used to build a dynamic heat balance model and a dynamic load assessment model based on multivariate fusion state data, and calculate the real-time operation state and safety margin of the power grid based on the dynamic capacity assessment model;

[0052] The model building module is used to integrate the measurement data of the nodes on the transmission line at the past L time points to obtain historical fused data. The historical fused data, the real-time operation status of the power grid, and the safety margin are input into the intelligent decision-making algorithm. The intelligent decision-making prediction model is constructed using the historical real load and renewable energy power generation as labels.

[0053] The strategy output module is used to integrate the measurement data of nodes on the real-time transmission line into a multi-dimensional form, combine it with the current real-time operating status and safety margin of the current power grid, and input it into the constructed intelligent decision-making prediction model to predict the power grid load and renewable energy power generation. Based on the power grid load and renewable energy power generation, a globally coordinated dynamic capacity expansion control strategy is generated to realize intelligent regulation of dynamic capacity expansion of transmission lines.

[0054] The third aspect of the present invention proposes an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor. When the computer program is loaded into the processor, the method for intelligently controlling dynamic capacity expansion of a transmission line according to the first aspect of the present invention is implemented.

[0055] A fourth aspect of the present invention provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, the method for intelligently controlling dynamic capacity expansion of a transmission line according to the first aspect of the present invention is implemented.

[0056] Compared with the existing technology, the present invention has at least the following beneficial effects: without changing the original grid operating conditions, the present invention uses multi-source data fusion technology to unify the modeling of renewable energy generation, environment, load and equipment status data. Through methods such as Kalman filtering and Bayesian networks, data noise reduction and fusion are achieved, and grid data in complex scenarios is quantified and unified, thereby improving the accuracy of the description of the actual operating status of the current grid.

[0057] The present invention constructs a dynamic thermal balance model, associates environmental parameters with conductor current carrying capacity, and dynamically calculates the grid capacity limit, namely the safety margin solved by the present invention. This overcomes the limitation of existing technologies that rely on conservative and fixed safety thresholds to set capacity expansion space, resulting in inaccurate capacity expansion range and thus grid risks. By dynamically adjusting the safety margin through data integration, the capacity expansion space during renewable energy fluctuations is increased, ensuring that under conditions of renewable energy fluctuations and load changes, the grid can both operate stably and fully tap its potential capacity expansion capacity, thereby improving overall resource utilization and the level of renewable energy absorption. When environmental conditions are favorable, such as but not limited to low temperatures, high wind speeds, and weak sunlight, the actual transmission capacity of the grid can be significantly improved.

[0058] The present invention predicts load and renewable energy power generation through an intelligent decision-making algorithm inputted with safety margin and real-time status of the power grid, thereby improving the accuracy of power grid prediction, dynamically increasing capacity based on load and renewable energy power generation, improving the safety of dynamic capacity expansion of the power grid, improving the efficiency of responding to renewable energy fluctuations, and reducing the operation risks of the power grid. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1 It is a schematic diagram of the overall structure of a system provided according to an embodiment of the present invention. DETAILED DESCRIPTION

[0060] To make the objectives, technical solutions, and advantages of the present invention more clear, the technical solutions of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of the embodiments of the present invention. The described embodiments are only part of the embodiments of the present invention, not all of them. Based on the spirit of the present invention, all other embodiments obtained by ordinary technicians in this field without making any creative efforts are within the scope of protection of the present invention.

[0061] like Figure 1 As shown, embodiment 1 of the present invention provides a method for intelligent management of dynamic capacity expansion of a transmission line, comprising the following steps:

[0062] Step 1: Collect measurement data on new energy, environment, load and equipment status of nodes on the transmission line.

[0063] In a preferred but non-limiting embodiment of the present invention, step 1 comprises:

[0064] In a certain new energy power grid, new energy power generation monitoring devices, environmental monitoring devices, load monitoring devices and equipment status monitoring devices are installed at nodes on substations, distributed new energy stations and transmission lines in key load areas to collect real-time data such as solar energy, wind energy, temperature, humidity, wind speed, current, and load.

[0065] Further preferably, the new energy power generation monitoring device collects data such as the power generation power of the solar panels and wind turbine generator sets, the inverter output voltage and current;

[0066] Environmental monitoring devices collect meteorological data such as temperature, humidity, wind speed, and sunshine intensity in real time through sensors;

[0067] The load detection device monitors the real-time load data of each area to reflect the user's electricity consumption;

[0068] The equipment status monitoring device collects the operating status, vibration, temperature and other information of key equipment (such as circuit breakers, transformers, etc.).

[0069] Step 2: pre-process the measurement data collected from the nodes on the transmission line to obtain pre-processed measurement data.

[0070] In a preferred but non-limiting embodiment of the present invention, step 2 comprises:

[0071] The collected data is transmitted to the regional data collection terminal through wireless or optical fiber communication. The terminal filters, denoises, normalizes and timestamps the collected data to ensure data quality and timeliness.

[0072] Step 3: Use Kalman filtering, Bayesian network and multi-sensor fusion algorithm to fuse the pre-processed measurement data into multivariate fusion state data.

[0073] In a preferred but non-limiting embodiment of the present invention, step 3 comprises:

[0074] Step 3.1: Perform Kalman filter optimal state estimation on the preprocessed measurement data.

[0075] Further preferably, step 3.1 includes:

[0076] Step 3.1.1: Construct a state space model based on the preprocessed measurement data, which can be expressed as follows:

[0077]

[0078] Where,

[0079] x k is the implicit true state, x k-1 is the state at the previous moment,

[0080] F is the state transfer matrix,

[0081] H is the observation matrix,

[0082] w k 、v k are process noise and measurement noise, respectively.

[0083] z k represents the preprocessed measurement data, in, is the pre-processed measurement value of the i-th type monitoring device at time k.

[0084] Step 3.1.2: Perform Kalman filtering on the state space model of step 3.1.1 to obtain the optimal state estimate at this moment.

[0085] More preferably, step 3.1.2 comprises:

[0086] The Kalman filter prediction of the state space model is expressed as follows:

[0087]

[0088] P k|k-1 =FP k-1|k-1 F T +Q (3)

[0089] Where,

[0090] represents the state prediction value based on the kth moment,

[0091] represents the optimal estimated value of the state at the k-1th moment,

[0092] P k|k-1 represents the covariance matrix of the state prediction value at the kth moment,

[0093] P k-1|k-1 represents the covariance matrix of the optimal estimate of the state at the k-1th moment,

[0094] Q represents the process noise covariance matrix.

[0095] The prediction results are updated by Kalman filtering and expressed as follows:

[0096] K k =P k|k-1 H T [HP k|k-1 H T +R] -1 (2)

[0097]

[0098] P k|k =(IK k H)P k|k-1 (4)

[0099] Where,

[0100] K k represents the Kalman gain,

[0101] R represents the observation noise covariance matrix,

[0102] represents the optimal estimated value of the state at the kth moment,

[0103] P k|k represents the covariance matrix of the optimal estimate of the state at the kth moment,

[0104] I represents the identity matrix,

[0105] Step 3.2: Based on the optimal state estimate obtained in step 3.1, a Bayesian network is constructed and probabilistic inference is performed to obtain the posterior probability distribution of the Bayesian network output.

[0106] Further preferably, step 3.2 includes:

[0107] The optimal state estimate output by the Kalman filter Input the Bayesian network, and then use the Bayesian formula for probability reasoning to solve the posterior probability distribution The posterior probability distribution is expressed as follows:

[0108]

[0109] in,

[0110] Indicates Under the condition of , the probability distribution of the hidden variable set Θ in the Bayesian network is,

[0111] When the latent variable takes the value Θ, we observe The probability of

[0112] express The probability of occurrence is expressed as follows:

[0113]

[0114] Posterior probability distribution Expressed as the set of posterior distributions of each latent variable n is the total number of latent variables, where Θ represents each latent variable in the unified model. The Bayesian network structure is used to define each latent variable corresponding to the node set {Θ1,…,Θ n}, and specify the conditional probability table P(Θ i ∣Pa(Θ i )), represents the hidden variable Θ corresponding to each node set i At its parent node Pa(Θ i ) is the conditional probability distribution when the value is determined.

[0115] In step 3.3, based on the posterior probability distribution of step 3.2, the maximum a posteriori estimate is calculated to generate the multivariate fusion state data, which is expressed as follows:

[0116]

[0117] Where,

[0118] Represents multivariate fusion status data.

[0119] Step 4: Based on the multivariate fusion state data in step 3, a dynamic heat balance model and a dynamic load assessment model are constructed, and the real-time operation state of the power grid is calculated based on the dynamic capacity assessment model, and the real-time operation state and safety margin of the current power grid are calculated.

[0120] In a preferred but non-limiting embodiment of the present invention, step 4 comprises:

[0121] Step 4.1: Based on the multivariate fusion state data of step 3, a dynamic heat balance model is constructed to obtain the solar heat absorption term, convective heat dissipation term, radiative heat dissipation term, and evaporative cooling term.

[0122] Construct a dynamic heat balance model, which is expressed as the following formula:

[0123]

[0124] in,

[0125] mC p is the heat capacity per unit length of wire;

[0126] represents the rate of temperature change,

[0127] T(k) is the conductor temperature in the multivariate fusion state data,

[0128] I(k) is the load current in the multivariate fusion state data;

[0129] R(T) is the temperature resistance;

[0130] Q solar (k) is the solar heat absorption term;

[0131] Q conv (k) is the convective heat dissipation term;

[0132] Q rad (k) is the radiation heat dissipation term;

[0133] Q cvap (k) is the evaporative cooling term.

[0134] In step 4.2, a load assessment model is constructed based on the solar heat absorption term, convective heat dissipation term, radiant heat dissipation term, and evaporative cooling term obtained in step 4.1 to solve the current maximum safe current carrying capacity. The dynamic load assessment model is constructed using the current maximum safe current carrying capacity as the upper limit of the dynamic load current.

[0135] Further preferably, step 4.2 includes:

[0136] Step 4.2.1: Sum the solar heat absorption term, convection heat dissipation term, radiation heat dissipation term, and evaporative cooling term obtained in step 4.1, square the summation result with the ratio of the resistance at the maximum allowable temperature, construct a load assessment model, and solve the load assessment model to obtain the current maximum safe current carrying capacity I max , expressed as follows:

[0137]

[0138] Where,

[0139] R(T max ) represents the maximum allowable temperature T max The resistance when

[0140] Step 4.2.2: Subtract the theoretical maximum load of the line from the real-time load data, and multiply the subtraction result by the load margin coefficient to obtain the load multiplication result;

[0141] Multiply the real-time renewable energy power generation data by the renewable energy power generation contribution weight to obtain the renewable energy power generation multiplication result;

[0142] Multiply the real-time device status data by the device status coefficient to obtain the device status multiplication result;

[0143] Add the load multiplication result and the new energy power generation multiplication result, and then subtract the equipment status multiplication result to obtain the dynamic load current. Set the upper limit of the dynamic load current in step 4.2.1 to the current maximum safe current carrying capacity, and construct a dynamic load assessment model as shown in the following formula:

[0144] C dynamic (t) = α·E new (t)+β·(L max -L(t))-γ·D(t) (12)

[0145] in,

[0146] C dynamic (t) represents the dynamic load current, the upper limit is I max ,

[0147] E new (t) is the real-time new energy power generation data,

[0148] L max is the maximum load that the line can theoretically bear,

[0149] L(t) is the real-time load data,

[0150] D(t) is the real-time device status data,

[0151] α, β and γ are the contribution weights of renewable energy power generation, load margin coefficient and equipment status coefficient, which are all empirical weights and can be adjusted according to actual conditions.

[0152] Step 4.3: Calculate the real-time operation status of the power grid based on the dynamic load assessment model, and solve the current real-time operation status and safety margin of the power grid.

[0153] Further preferably, step 4.3 includes:

[0154] The dynamic load current obtained by solving the dynamic load assessment model, the conductor temperature in the multivariate fusion state data, and the load current in the multivariate fusion state data are combined to obtain the real-time operation status of the power grid;

[0155] The dynamic load current obtained by solving the dynamic load assessment model is subtracted from the load current in the multivariate fusion state data to obtain the safety margin, which is expressed as the following formula:

[0156] S(k)=[T(k),C dynamic (k),I(k)] T (13)

[0157] M(k)=C dynamic (k)-I(k) (14)

[0158] in,

[0159] S(k) represents the real-time operating status of the power grid,

[0160] M(k) represents the safety margin.

[0161] In step 5, the measurement data of the nodes on the transmission line at the past L time points are multi-element fused to obtain historical fused data. The historical fused data and the real-time operation status and safety margin of the power grid obtained in step 4 are input into the intelligent decision-making algorithm. The historical real load and renewable energy power generation are used as labels to build an intelligent decision-making prediction model.

[0162] In a preferred but non-limiting embodiment of the present invention, step 5 comprises:

[0163] Step 5.1: The measurement data of the nodes on the transmission line at the past L time points are multivariately integrated to obtain historical fusion data. The historical fusion data and the real-time operation status and safety margin of the power grid obtained in step 4 are combined into a decision feature vector:

[0164]

[0165] In step 5.2, the decision feature vector of step 5.1 is output as the load forecast and new energy power generation forecast at the next moment through the intelligent decision-making algorithm to construct an intelligent decision-making prediction model, wherein the intelligent decision-making algorithm includes a convolutional neural network, a long-short-term neural network and a support vector machine.

[0166] Further preferably, step 5.2 includes:

[0167] Step 5.2.1: Input the decision feature vector of step 5.1 into the convolutional neural network CNN(·) in the intelligent decision prediction model and extract the spatiotemporal features through the one-dimensional convolution layer to obtain the convolution output X CNN , expressed as follows:

[0168] X CNN =CNN(X k ) (16)

[0169] Step 5.2.2, convolution output XCNN , extracting the hidden state X from the spatiotemporal features through the long short-term neural network LSTM(·) in the intelligent decision prediction model LSTM , expressed as follows:

[0170] X LSTM =LSTM(X CNN ) (17)

[0171] Step 5.2.3, compare the hidden state with the environmental data X of the transmission line node at this moment env Splicing, then inputting the support vector machine (SVM) in the intelligent decision-making prediction model to make residual corrections to the splicing results and output the load forecast for the next moment. and new energy power generation forecast It is expressed as the following formula:

[0172]

[0173] Step 5.3, based on the current real load and renewable energy power generation The load forecast error and renewable energy generation error are solved for the label of the intelligent decision-making prediction model and the prediction result of step 5.2. When the load forecast error and renewable energy generation error are both greater than the threshold, the stochastic gradient descent method is used to update the parameters of the intelligent decision-making prediction model to achieve continuous adaptive update of the prediction model and adaptively update the parameters until the load forecast error and renewable energy generation error are both less than the set threshold.

[0174] Further preferably, step 5.3 includes:

[0175] Step 5.3.1: Using the current actual load and renewable energy power generation as labels for the intelligent decision-making prediction model, the load prediction error is calculated by subtracting the current actual load from the load prediction result, and the renewable energy power generation error is calculated by subtracting the current actual renewable energy power generation from the renewable energy power generation prediction result.

[0176] Compare the load forecast error with the load error threshold to obtain the load comparison result, and compare the renewable energy generation error with the renewable energy generation threshold to obtain the renewable energy comparison result, which is expressed as the following formula:

[0177]

[0178] Where,

[0179] is the actual load value, is the predicted load value,

[0180] is the load forecast error, thres1 is the load error threshold,

[0181] e(k+1) is the actual renewable energy power generation, To predict the amount of electricity generated by renewable energy,

[0182] εe(k+1) is the renewable energy power generation error, and thres2 is the renewable energy power generation threshold.

[0183] Step 5.3.2: When the load comparison result and the new energy comparison result are greater than the set load error threshold and new energy generation threshold, respectively, the parameters of the intelligent decision-making prediction model are updated using the stochastic gradient descent method;

[0184] The iteration is continued until the load forecast error and the renewable energy power generation error are both smaller than the set load error threshold and renewable energy power generation threshold respectively. The iteration is terminated and the training of the intelligent decision-making prediction model is completed.

[0185] More preferably, step 5.3.2 comprises:

[0186] Calculate the mean square error loss based on the load forecast error and the renewable energy generation error;

[0187] The stochastic gradient descent method is used to calculate the gradient of the mean square error loss with respect to the parameters in the intelligent decision prediction model. The parameters include the convolution kernel weights and bias of the convolutional neural network, the weight matrix and hidden state of the long short-term memory network, and the weight vector and intercept of the support vector machine.

[0188] The parameters are updated accordingly according to the gradient until the load forecast error and the renewable energy power generation error are both less than the set threshold, and the iteration terminates.

[0189] In step 6, the measurement data of the nodes on the real-time transmission line are multi-element fused, combined with the real-time operating status and safety margin of the power grid, and input into the intelligent decision-making prediction model constructed in step 5 to predict the power grid load and renewable energy power generation. Based on the power grid load and renewable energy power generation, a global coordinated dynamic capacity expansion control strategy is generated to realize intelligent regulation of dynamic capacity expansion of transmission lines.

[0190] In a preferred but non-limiting embodiment of the present invention, step 6 comprises:

[0191] All collected real-time and historical data is stored in a distributed database. Periodically analyzing historical data, the system optimizes data fusion models and intelligent decision-making algorithms. Leveraging machine learning techniques to continuously update prediction models, the system automatically generates load dispatch recommendations and provides decision support to the dispatch center, enabling continuous optimization of grid operations. Based on the intelligent decision-making results, control strategies are automatically generated, including adjusting grid operating parameters, activating auxiliary regulation devices, optimizing inverter settings, and issuing load regulation recommendations.

[0192] If the difference between the predicted value and the dynamic capacity is greater than zero, additional load or renewable energy can be freely allocated; otherwise, load reduction recommendations or standby unit activation are triggered. Dispatch recommendations are converted into specific unit output, energy storage charging and discharging, or controllable load response quantities using a built-in optimal power flow or pre-trained lightweight quadratic programming solver. Based on the difference between the predicted value and the dynamic capacity, grid operating parameters are adjusted to adjust transformer tap positions, altering the bus voltage gradient. Auxiliary regulation devices activate air cooling spray or conductor surface cooling when temperature or sag exceeds limits. Inverter settings reset the upper and lower limits of the injected power of distributed photovoltaic / wind power inverters in real time to ensure bus voltage and power flow balance. These steps are automatically coordinated via a unified control bus, forming a coupled closed loop of "prediction → mapping → distribution → feedback → fine-tuning." The collaborative decision-making module issues control commands to each terminal via a high-speed communication network. Upon receiving the commands, the terminal device automatically controls the on-site equipment to adjust parameters and optimize its status.

[0193] Only by generating a global dynamic capacity expansion control strategy and coordinating the control instructions of each terminal through a distributed consensus mechanism, can the control commands be finally sent to the on-site equipment to complete the grid parameter adjustment and load regulation, ensuring the safe and efficient operation of the grid.

[0194] Embodiment 2 of the present invention provides a transmission line dynamic capacity expansion intelligent control system, which runs the transmission line dynamic capacity expansion intelligent control method described in embodiment 1, including:

[0195] Data acquisition module, used to collect measurement data on new energy, environment, load and equipment status of nodes on the transmission line;

[0196] A preprocessing module is used to preprocess the measurement data collected from the nodes on the transmission line to obtain preprocessed measurement data;

[0197] Data fusion module, used to fuse pre-processed measurement data into multivariate fusion state data using Kalman filtering, Bayesian network and multi-sensor fusion algorithm;

[0198] The real-time state solving module of the power grid is used to build a dynamic heat balance model and a dynamic load assessment model based on multivariate fusion state data, and calculate the real-time operation state and safety margin of the power grid based on the dynamic capacity assessment model;

[0199] The model building module is used to integrate the measurement data of the nodes on the transmission line at the past L time points to obtain historical fused data. The historical fused data, the real-time operation status of the power grid, and the safety margin are input into the intelligent decision-making algorithm. The intelligent decision-making prediction model is constructed using the historical real load and renewable energy power generation as labels.

[0200] The strategy output module is used to integrate the measurement data of nodes on the real-time transmission line into a multi-dimensional form, combine it with the current real-time operating status and safety margin of the current power grid, and input it into the constructed intelligent decision-making prediction model to predict the power grid load and renewable energy power generation. Based on the power grid load and renewable energy power generation, a globally coordinated dynamic capacity expansion control strategy is generated to realize intelligent regulation of dynamic capacity expansion of transmission lines.

[0201] Embodiment 3 of the present invention provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the computer program is loaded into the processor, the method for intelligently controlling dynamic capacity expansion of a transmission line according to embodiment 1 is implemented.

[0202] Embodiment 4 of the present invention provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, it implements the intelligent control method for dynamic capacity expansion of a transmission line according to embodiment 1.

[0203] Compared with the existing technology, the present invention has at least the following beneficial effects: without changing the original grid operating conditions, the present invention uses multi-source data fusion technology to unify the modeling of renewable energy generation, environment, load and equipment status data. Through methods such as Kalman filtering and Bayesian networks, data noise reduction and fusion are achieved, and grid data in complex scenarios is quantified and unified, thereby improving the accuracy of the description of the actual operating status of the current grid.

[0204] The present invention constructs a dynamic thermal balance model, associates environmental parameters with conductor current carrying capacity, and dynamically calculates the grid capacity limit, namely the safety margin solved by the present invention. This overcomes the limitation of existing technologies that rely on conservative and fixed safety thresholds to set capacity expansion space, resulting in inaccurate capacity expansion range and thus grid risks. By dynamically adjusting the safety margin through data integration, the capacity expansion space during renewable energy fluctuations is increased, ensuring that under conditions of renewable energy fluctuations and load changes, the grid can both operate stably and fully tap its potential capacity expansion capacity, thereby improving overall resource utilization and the level of renewable energy absorption. When environmental conditions are favorable, such as but not limited to low temperatures, high wind speeds, and weak sunlight, the actual transmission capacity of the grid can be significantly improved.

[0205] The present invention predicts load and renewable energy power generation through an intelligent decision-making algorithm inputted with safety margin and real-time status of the power grid, thereby improving the accuracy of power grid prediction, dynamically increasing capacity based on load and renewable energy power generation, improving the safety of dynamic capacity expansion of the power grid, improving the efficiency of responding to renewable energy fluctuations, and reducing the operation risks of the power grid.

[0206] The present disclosure may be a system, method and / or computer program product. The computer program product may include a computer-readable storage medium carrying computer-readable program instructions for causing a processor to implement various aspects of the present disclosure.

[0207] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered by the scope of protection of the claims of the present invention.

Claims

1. A method for intelligently controlling dynamic capacity expansion of a transmission line, characterized by: Collect measurement data on new energy, environment, load and equipment status at nodes on the transmission line; Preprocessing the measurement data collected from nodes on the transmission line to obtain preprocessed measurement data; Using Kalman filtering, Bayesian network and multi-sensor fusion algorithm, the pre-processed measurement data is fused into multivariate fusion state data; Based on the multivariate fusion state data, a dynamic heat balance model and a dynamic load assessment model are constructed, and the real-time operation status and safety margin of the power grid are calculated based on the dynamic capacity assessment model; The measurement data of the nodes on the transmission line at the past L time points are multi-element fused to obtain historical fused data. The historical fused data, the real-time operation status of the power grid, and the safety margin are input into the intelligent decision-making algorithm. The intelligent decision-making prediction model is constructed using the historical real load and renewable energy power generation as labels. After multi-dimensional fusion of the measurement data of nodes on the real-time transmission line, it is combined with the current real-time operating status and safety margin of the current power grid, and input into the constructed intelligent decision-making prediction model to predict the power grid load and renewable energy power generation, and generate a globally coordinated dynamic capacity expansion control strategy based on the power grid load and renewable energy power generation.

2. The method for intelligently controlling dynamic capacity expansion of a power transmission line according to claim 1, characterized in that: The method of using Kalman filtering, Bayesian network and multi-sensor fusion algorithm to fuse pre-processed measurement data into multivariate fusion state data specifically includes: Perform Kalman filter optimal state estimation on pre-processed measurement data; Based on the optimal state estimation, a Bayesian network is constructed and probabilistic reasoning is performed to obtain the posterior probability distribution of the Bayesian network output; According to the posterior probability distribution, the maximum a posteriori estimate is calculated to generate multivariate fusion state data.

3. The method for intelligently controlling dynamic capacity expansion of a power transmission line according to claim 1, characterized in that: The method of constructing a dynamic heat balance model and a dynamic load assessment model based on the multivariate fusion state data and calculating the real-time operation state and safety margin of the power grid based on the dynamic capacity assessment model specifically includes: Based on the multivariate fusion state data, a dynamic heat balance model is constructed to obtain the solar heat absorption term, convective heat dissipation term, radiative heat dissipation term and evaporative cooling term; Based on the obtained solar heat absorption term, convection heat dissipation term, radiation heat dissipation term and evaporative cooling term, a load assessment model is constructed to solve the current maximum safe current carrying capacity. The current maximum safe current carrying capacity is used as the upper limit of the dynamic load current to construct a dynamic load assessment model. The real-time operation status of the power grid is calculated based on the dynamic load assessment model, and the current real-time operation status and safety margin of the power grid are solved.

4. The method for intelligently controlling dynamic capacity expansion of a power transmission line according to claim 3, characterized in that: The load assessment model is constructed based on the obtained solar heat absorption term, convection heat dissipation term, radiation heat dissipation term and evaporative cooling term, and the current maximum safe current carrying capacity is solved. The dynamic load assessment model is constructed with the current maximum safe current carrying capacity as the upper limit of the dynamic load current, specifically including: The solar heat absorption term, convection heat dissipation term, radiation heat dissipation term, and evaporative cooling term are summed, and the square root of the summation result and the ratio of the resistance at the maximum allowable temperature are taken to construct a load assessment model. The load assessment model is solved to obtain the current maximum safe current carrying capacity. Subtract the theoretical maximum load of the line from the real-time load data, and multiply the subtraction result by the load margin coefficient to obtain the load multiplication result; Multiply the real-time renewable energy power generation data by the renewable energy power generation contribution weight to obtain the renewable energy power generation multiplication result; Multiply the real-time device status data by the device status coefficient to obtain the device status multiplication result; The dynamic load current is obtained by adding the load multiplication result and the new energy power generation multiplication result, and then subtracting the equipment status multiplication result. The upper limit of the dynamic load current is set to the current maximum safe current carrying capacity to construct a dynamic load assessment model.

5. The method for intelligently controlling dynamic capacity expansion of a power transmission line according to claim 3, characterized in that: The calculation of the real-time operation status of the power grid based on the dynamic load assessment model and the solution of the current real-time operation status and safety margin of the power grid specifically include: The dynamic load current obtained by solving the dynamic load assessment model, the conductor temperature in the multivariate fusion state data, and the load current in the multivariate fusion state data are combined to obtain the real-time operation status of the power grid; The dynamic load current obtained by solving the dynamic load assessment model is subtracted from the load current in the multivariate fusion state data to obtain the safety margin.

6. The method for intelligently controlling dynamic capacity expansion of a power transmission line according to claim 1, characterized in that: The method involves fusing the measurement data of the nodes on the transmission line at the past L time points to obtain historical fused data, inputting the historical fused data, the real-time operation status of the power grid, and the safety margin into the intelligent decision-making algorithm, and constructing an intelligent decision-making prediction model using the historical real load and renewable energy power generation as labels. Specifically, the method includes: The measurement data of the nodes on the transmission line at the past L time points are multivariately integrated to obtain historical fusion data, and the historical fusion data, the real-time operation status of the power grid and the safety margin are combined to obtain the decision feature vector; The decision feature vector is input into the intelligent decision algorithm in the intelligent decision prediction model to output the load forecast and renewable energy power generation forecast results at the next moment. The intelligent decision algorithm includes convolutional neural networks, long-short-term neural networks, and support vector machines. The current real load and renewable energy power generation are used as labels for the intelligent decision-making prediction model. The load forecast error and renewable energy power generation error are solved through the labels and prediction results. When the load forecast error and renewable energy power generation error are both greater than the set threshold, the stochastic gradient descent method is used to update the parameters of the intelligent decision-making prediction model. The parameters are adaptively updated until the load forecast error and renewable energy power generation error are both less than the set threshold, and the training of the intelligent decision-making prediction model is completed.

7. The method for intelligently controlling dynamic capacity expansion of a power transmission line according to claim 6, characterized in that: The intelligent decision algorithm inputting the decision feature vector into the intelligent decision prediction model outputs the load forecast and new energy power generation forecast results at the next moment, specifically including: The decision feature vector is input into the convolutional neural network in the intelligent decision prediction model, and the spatiotemporal feature is extracted through the one-dimensional convolution layer to obtain the convolution output; The convolution output is passed through the long-term and short-term neural network in the intelligent decision-making prediction model to extract the hidden state from the spatiotemporal features; The hidden state is spliced ​​with the environmental data of the transmission line nodes at this moment, and then input into the support vector machine in the intelligent decision-making prediction model to perform residual correction on the splicing result, and output the load forecast and renewable energy power generation forecast results at the next moment.

8. The method for intelligently controlling dynamic capacity expansion of a power transmission line according to claim 6, characterized in that: The current real load and renewable energy power generation are used as labels for the intelligent decision-making prediction model. The load forecast error and renewable energy power generation error are solved by using the labels and prediction results. When the load forecast error and renewable energy power generation error are both greater than a set threshold, the parameters of the intelligent decision-making prediction model are updated using the stochastic gradient descent method. The parameters are adaptively updated until the load forecast error and renewable energy power generation error are both less than the set threshold. The training of the intelligent decision-making prediction model is completed. Specifically, the following steps are performed: The current real load and renewable energy power generation are used as labels for the intelligent decision-making prediction model. The load prediction error is solved by taking the difference between the current real load and the load prediction result, and the renewable energy power generation error is solved by taking the difference between the current real renewable energy power generation and the renewable energy power generation prediction result. Compare the load forecast error with the load error threshold to obtain a load comparison result, and compare the new energy generation error with the new energy generation threshold to obtain a new energy comparison result; When the load comparison result and the new energy comparison result are respectively greater than the set load error threshold and new energy generation threshold, the parameters of the intelligent decision-making prediction model are updated using the stochastic gradient descent method; The iteration is continued until the load forecast error and the renewable energy power generation error are both smaller than the set load error threshold and renewable energy power generation threshold respectively. The iteration is terminated and the training of the intelligent decision-making prediction model is completed.

9. The method for intelligently controlling dynamic capacity expansion of a power transmission line according to claim 8, characterized in that: The method of updating the parameters of the intelligent decision-making prediction model using the stochastic gradient descent method specifically includes: Calculate the mean square error loss based on the load forecast error and the renewable energy generation error; The stochastic gradient descent method is used to calculate the gradient of the mean square error loss with respect to the parameters in the intelligent decision-making prediction model. The parameters are updated accordingly based on the gradient until the load forecast error and the renewable energy power generation error are both less than the set threshold. The iteration is terminated and the training of the intelligent decision-making prediction model is completed.

10. A transmission line dynamic capacity expansion intelligent control system, which runs a transmission line dynamic capacity expansion intelligent control method according to any one of claims 1 to 9, characterized in that: Data acquisition module, used to collect measurement data on new energy, environment, load and equipment status of nodes on the transmission line; A preprocessing module is used to preprocess the measurement data collected from the nodes on the transmission line to obtain preprocessed measurement data; Data fusion module, used to fuse pre-processed measurement data into multivariate fusion state data using Kalman filtering, Bayesian network and multi-sensor fusion algorithm; The real-time state solving module of the power grid is used to build a dynamic heat balance model and a dynamic load assessment model based on multivariate fusion state data, and calculate the real-time operation state and safety margin of the power grid based on the dynamic capacity assessment model; The model building module is used to integrate the measurement data of the nodes on the transmission line at the past L time points to obtain historical fused data. The historical fused data, the real-time operation status of the power grid, and the safety margin are input into the intelligent decision-making algorithm. The intelligent decision-making prediction model is constructed using the historical real load and renewable energy power generation as labels. The strategy output module is used to integrate the measurement data of nodes on the real-time transmission line into a multi-dimensional form, combine it with the current real-time operating status and safety margin of the current power grid, and input it into the constructed intelligent decision-making prediction model to predict the power grid load and renewable energy power generation. Based on the power grid load and renewable energy power generation, a globally coordinated dynamic capacity expansion control strategy is generated.

11. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the computer program is loaded into a processor, a method for intelligently controlling dynamic capacity increase of a power transmission line is implemented according to any one of claims 1 to 9.

12. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, a method for intelligently controlling dynamic capacity increase of a power transmission line is implemented according to any one of claims 1 to 9.

Images