Flywheel energy storage management system for renewable energy source integration

Abstract

The invention relates to the technical field of flywheel energy storage, in particular to a flywheel energy storage management system for renewable energy integration, which comprises a state monitoring module, an energy access module, a material fatigue prediction and optimization module and an energy scheduling module, the state monitoring module collects the operation state of the flywheel energy storage unit in real time. The energy access module receives input power from renewable energy sources; the material fatigue prediction and optimization module generates an operation optimization strategy when a fatigue risk is predicted; and the energy scheduling module performs dynamic energy distribution among the flywheel energy storage unit, the power grid and the standby energy storage unit, and formulates an optimal energy flow strategy in combination with historical load data and real-time input power. According to the invention, the service life of the flywheel energy storage unit is greatly prolonged, and the overall safety and reliability of the system are improved.

Description

Technical Field

[0001] The invention relates to the technical field of flywheel energy storage, and in particular to a flywheel energy storage management system for renewable energy integration. Background Art

[0002] As renewable energy (such as solar energy, wind energy, etc.) gradually becomes an important part of the global energy structure transformation, its volatility and instability problems are becoming increasingly prominent. In order to effectively utilize renewable energy and solve the uncertainty of energy supply in the power system, energy storage technology has become a key solution. Flywheel energy storage technology, as an efficient, reliable and fast-response energy storage method, has been widely used in many fields. Flywheel energy storage systems can store electrical energy through mechanical energy, have high energy conversion efficiency and long service life, and are particularly suitable for combination with renewable energy systems to provide power balance and stability. However, flywheel energy storage systems will face problems of material fatigue and damage during long-term operation, which requires real-time monitoring and optimization management to ensure their safety and efficiency.

[0003] Existing flywheel energy storage management systems mostly rely on simple monitoring technology and regular maintenance, which makes it difficult to accurately predict the fatigue degree and damage of flywheel materials in real time. The monitoring methods of existing technologies are usually limited to the collection of basic parameters such as the speed and temperature of the flywheel, and lack in-depth analysis of the fatigue state of the material, which makes the flywheel energy storage unit prone to safety hazards in long-term operation. In addition, traditional energy scheduling methods often ignore the health status of the energy storage unit, material fatigue and the volatility of different renewable energy inputs, resulting in low energy distribution efficiency and the stability and reliability of the system cannot be effectively guaranteed. Therefore, the existing technology has failed to make full use of advanced data analysis methods to optimize the operation and energy scheduling strategies of the flywheel energy storage system, and cannot maximize the operating efficiency and sustainability of the system. Summary of the invention

[0004] In view of the deficiencies in the prior art, the purpose of the present invention is to provide a flywheel energy storage management system for renewable energy integration, so as to achieve efficient energy scheduling, improve the stability, reliability and sustainability of the system, and extend the service life of the energy storage unit.

[0005] The invention provides a flywheel energy storage management system for renewable energy integration.

[0006] A flywheel energy storage management system for renewable energy integration includes a condition monitoring module, an energy access module, a material fatigue prediction and optimization module, and an energy scheduling module, wherein;

[0007] The state monitoring module collects the operating state of the flywheel energy storage unit in real time, including speed, temperature, vibration and material fatigue parameters;

[0008] The energy access module receives input power from renewable energy (solar energy, wind energy), and dynamically adjusts the input power from different sources using a real-time power matching algorithm;

[0009] The material fatigue prediction and optimization module predicts the fatigue degree of the flywheel material based on the real-time collected operating status of the flywheel energy storage unit using a convolutional neural network (CNN) model, and generates an operation optimization strategy when fatigue risk is predicted in combination with a fatigue cumulative damage model, specifically including:

[0010] Operation data preprocessing: preprocess the collected operation status of the flywheel energy storage unit, including data cleaning and standardization;

[0011] Material fatigue prediction: Based on the pre-processed operating status, the fatigue status of the flywheel material is evaluated in real time through the convolutional neural network (CNN) model, and the future fatigue trend is predicted;

[0012] Operation optimization strategy generation: Based on the results of material fatigue prediction and combined with the fatigue cumulative damage model, a dynamic operation optimization strategy is generated, including flywheel speed adjustment and load distribution optimization;

[0013] The energy scheduling module performs dynamic energy distribution among the flywheel energy storage unit, the power grid and the backup energy storage unit based on the generated operation optimization strategy and the operation status of the flywheel energy storage unit, and formulates the optimal energy flow strategy in combination with historical load data and real-time input power.

[0014] Optionally, the status monitoring module includes:

[0015] Speed ​​acquisition: The speed sensor (Hall effect sensor) is used to measure the speed of the flywheel in real time;

[0016] Temperature collection: Use temperature sensors to monitor temperature changes in the flywheel energy storage unit in real time, including motor temperature, bearing temperature, and flywheel internal temperature;

[0017] Vibration collection: Real-time monitoring of the vibration state of the flywheel energy storage unit through vibration sensors;

[0018] Material fatigue parameter collection: The strain and fatigue parameters of the flywheel energy storage unit material are monitored in real time through strain sensors.

[0019] Optionally, the energy access module includes:

[0020] Input power reception: receiving power input from solar energy and wind energy through the power access unit;

[0021] Real-time power matching: Based on the real-time data of power input, a real-time power matching algorithm is used for dynamic adjustment.

[0022] Optionally, the operation data preprocessing includes:

[0023] Data cleaning: Clean the collected operating status data of the flywheel energy storage unit to remove noise, outliers and missing data;

[0024] Standardization processing: The operating status data after cleaning is standardized to eliminate the dimensional differences between different parameters.

[0025] Optionally, the convolutional neural network (CNN) model includes:

[0026] Input data processing: Perform time series conversion on the pre-processed operating status data (including speed, temperature, vibration, and material fatigue parameters) and convert the operating status data x t Convert to time series data X = {x1, x2, ..., x N}, where x1,x2,...,x N are the flywheel state features collected at time 1, 2, ..., N, where N is the size of the time window;

[0027] Convolutional layer design: By using two convolution kernels K1 and K2, with sizes of 1×3, 1×5, 1×7 and 1×2, 1×4, 1×6 respectively, feature maps Y1 and Y2 at different scales are extracted;

[0028] Feature map merging: Merge the feature maps Y1 and Y2 obtained from different convolution kernels to form the final feature map Y;

[0029] Pooling operation: After the output of the convolutional layer, the final feature map is downsampled by using a 2×2 maximum pooling window;

[0030] Fully connected layer and fatigue state prediction: The feature map obtained after convolution and pooling is input into the fully connected layer to predict the material fatigue state;

[0031] Fatigue trend prediction: Based on the results of the material fatigue state prediction, a recurrent neural network (RNN) is used to predict the future fatigue state.

[0032] Optionally, the operation optimization strategy generation includes:

[0033] Calculation of damage degree of flywheel material: Based on the fatigue prediction results of flywheel material, the fatigue cumulative damage model is used to calculate the damage degree D(t) of flywheel material;

[0034] Fatigue risk assessment: Based on the calculated damage degree D(t) of the flywheel material, the fatigue risk R(t) is assessed;

[0035] Operation optimization strategy generation: When the fatigue risk R(t) reaches or exceeds the preset fatigue threshold R th When , the operation optimization strategy is started, including adjusting the flywheel speed ω(t) and load distribution P(t).

[0036] Optionally, the energy scheduling module includes:

[0037] Energy allocation strategy generation: Based on the generated operation optimization strategy (including flywheel speed adjustment and load distribution optimization), combined with the real-time operation status of the flywheel energy storage unit, power grid, and backup energy storage unit, the current power demand and the charging and discharging status of the energy storage unit are dynamically evaluated to generate the optimal energy allocation strategy;

[0038] Energy flow regulation: According to the generated energy distribution strategy, the energy flow between the flywheel energy storage unit, the power grid and the backup energy storage unit is adjusted in real time;

[0039] Load forecasting and historical data optimization: Combine historical load data with real-time input power to predict future load demand and optimize energy scheduling strategies.

[0040] Optionally, the energy allocation strategy generation includes:

[0041] Power demand assessment: Combine the real-time operating status of the flywheel energy storage unit, the power grid and the backup energy storage unit to dynamically assess the current power demand and calculate the required power P demand (t);

[0042] Optimal energy allocation strategy generation: Based on the dynamically evaluated power demand and the state of the energy storage unit, an optimization algorithm is used to generate the optimal energy allocation strategy S optimal (t), including the energy flow between the flywheel energy storage unit, the grid and the backup energy storage unit;

[0043] Generate strategy output: The final output energy allocation strategy S optimal (t) is used to adjust the energy flow between the flywheel energy storage unit, the power grid and the backup energy storage unit.

[0044] Optionally, the energy flow regulation includes:

[0045] Real-time power output calculation: According to the generated energy allocation strategy, the power output value of the flywheel energy storage unit is calculated, which is expressed as:

[0046] P flywheel (t) = P opt (t)-P grid (t)-P backup (t);

[0047] Among them, P flywheel(t) is the power output of the flywheel energy storage unit at time t, P opt (t) is the optimized power generated according to the energy allocation strategy, P grid (t) is the power flowing into or out of the grid, P backup (t) is the power inflow or outflow of the backup energy storage unit;

[0048] Energy flow adjustment condition judgment: After calculating the power output of the flywheel energy storage unit, determine that the power output of the flywheel energy storage unit does not exceed its maximum and minimum power range. If the power output of the flywheel energy storage unit exceeds its rated power range, maintain power balance by adjusting the power inflow and outflow of the power grid and the backup energy storage unit.

[0049] Optionally, the load forecasting and historical data optimization includes:

[0050] Historical load data collection: Collect historical load data from the power grid and flywheel energy storage unit;

[0051] Load demand forecasting: Based on the collected historical load data, the ARIMA model is used to perform time series analysis to predict future load demand;

[0052] Optimize energy dispatch strategy: Based on the load demand forecast results, combined with real-time power input and energy storage status, use optimization algorithms to optimize energy flow and generate the best dispatch strategy;

[0053] Energy dispatch strategy implementation: According to the optimized energy dispatch strategy, the energy flow between the flywheel energy storage unit, the power grid and the backup energy storage unit is adjusted.

[0054] Beneficial effects of the present invention:

[0055] The present invention can monitor the operating status of the flywheel energy storage unit in real time, including rotation speed, temperature, vibration and material fatigue parameters. This real-time data acquisition and dynamic analysis capability enables the system to identify potential mechanical damage, overheating or material fatigue problems in advance, thereby reducing the risk of failure. The convolutional neural network model is used to accurately predict the fatigue state of the flywheel material, which greatly extends the service life of the flywheel energy storage unit and improves the overall safety and reliability of the system.

[0056] The present invention, through the energy access module, can dynamically adjust the input power from solar energy and wind energy to ensure the stability and efficiency of the flywheel energy storage unit during the charging process. The application of the real-time power matching algorithm optimizes the integration process of different energy inputs and effectively responds to the challenges brought by the volatility of renewable energy. In addition, the energy scheduling module generates the optimal energy flow strategy based on the operating status of the flywheel energy storage unit, the power grid and the backup energy storage unit, combined with historical load data and real-time power demand, ensuring the efficient allocation of energy and improving the energy utilization rate of the system.

[0057] The present invention, by comprehensively considering the health status of the flywheel energy storage unit, the material fatigue prediction results and the real-time load demand, can dynamically adjust the flywheel speed, load distribution and energy flow, and optimize the operation strategy of the entire system. When fatigue risk is detected, the system can generate an operation optimization strategy in time to ensure that the flywheel energy storage unit operates in the optimal state, reduce energy waste and improve the reliability of the system. At the same time, a load demand prediction method based on historical load data and real-time input power is adopted, which can predict future load demands in advance, further optimize the energy scheduling strategy, and ensure that the system maintains stability and sustainability in long-term operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are only for the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0059] Figure 1 A schematic diagram of system function modules according to an embodiment of the present invention;

[0060] Figure 2 Schematic diagram of a material fatigue prediction and optimization module according to an embodiment of the present invention. DETAILED DESCRIPTION

[0061] The present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments. At the same time, it is explained here that in order to make the embodiments more detailed, the following embodiments are the best and preferred embodiments, and those skilled in the art may also adopt other alternatives to implement some known technologies; and the accompanying drawings are only for more specific description of the embodiments, and are not intended to specifically limit the present invention.

[0062] like Figure 1-Figure 2 As shown, a flywheel energy storage management system for renewable energy integration includes a state monitoring module, an energy access module, a material fatigue prediction and optimization module, and an energy scheduling module, wherein;

[0063] The status monitoring module collects the operating status of the flywheel energy storage unit in real time, including speed, temperature, vibration and material fatigue parameters;

[0064] The energy access module receives input power from renewable energy (solar energy, wind energy), and uses a real-time power matching algorithm to dynamically adjust the input power from different sources to ensure the charging stability and efficiency of the flywheel energy storage unit;

[0065] The material fatigue prediction and optimization module uses the convolutional neural network (CNN) model to predict the fatigue degree of the flywheel material based on the real-time acquisition of the operating status of the flywheel energy storage unit, and combines the fatigue cumulative damage model to generate an operation optimization strategy when fatigue risk is predicted, including:

[0066] Operation data preprocessing: preprocess the collected operation status of the flywheel energy storage unit, including data cleaning and standardization;

[0067] Material fatigue prediction: Based on the pre-processed operating status, the fatigue status of the flywheel material is evaluated in real time through the convolutional neural network (CNN) model, and the future fatigue trend is predicted;

[0068] Operation optimization strategy generation: Based on the results of material fatigue prediction and combined with the fatigue cumulative damage model, a dynamic operation optimization strategy is generated, including flywheel speed adjustment and load distribution optimization;

[0069] The energy dispatch module dynamically distributes energy between the flywheel energy storage unit, the power grid and the backup energy storage unit based on the generated operation optimization strategy and the operating status of the flywheel energy storage unit, and formulates the optimal energy flow strategy based on historical load data and real-time input power;

[0070] Through the above content, the high efficiency and reliability of the system are achieved, which not only improves the service life of the flywheel energy storage unit, but also solves the safety hazard caused by material fatigue by real-time optimization of operating parameters. At the same time, it dynamically adapts to the fluctuating input of renewable energy, realizes efficient energy distribution, and improves the operating stability and sustainability of the overall system.

[0071] The condition monitoring module includes:

[0072] Speed ​​acquisition: The speed sensor (Hall effect sensor) is used to measure the speed of the flywheel in real time;

[0073] Temperature collection: Use temperature sensors to monitor temperature changes in the flywheel energy storage unit in real time, including motor temperature, bearing temperature, and flywheel internal temperature;

[0074] Vibration collection: Real-time monitoring of the vibration state of the flywheel energy storage unit through vibration sensors;

[0075] Material fatigue parameter collection: real-time monitoring of the strain and fatigue parameters of the flywheel energy storage unit material through strain sensors;

[0076] Through the above content, real-time monitoring and fault warning of the flywheel energy storage system can be achieved, and potential mechanical damage, overheating or material fatigue problems can be discovered in time, avoiding the problems of delayed detection and insufficient warning in traditional technologies. At the same time, accurate data collection provides a scientific basis for subsequent material fatigue prediction, operation optimization and energy scheduling, improves the safety, reliability and operation efficiency of the system, extends the service life of the flywheel energy storage unit, reduces maintenance costs, and enhances the system's adaptability to the volatility of renewable energy.

[0077] Energy access modules include:

[0078] Input power reception: receiving power input from solar energy and wind energy through the power access unit;

[0079] Real-time power matching: Based on the real-time data of power input, a real-time power matching algorithm is used for dynamic adjustment to ensure the charging stability of the flywheel energy storage unit, which is expressed as:

[0080] P adjusted =P solar +P wind -P loss ;

[0081] Among them, P adjusted is the adjusted total input power, P solar Power provided by solar panels, P wind Power provided to wind turbines, P loss is the system power loss;

[0082] Through the above content, it is possible to effectively integrate input power from different renewable energy sources (such as solar energy and wind energy) and make dynamic adjustments based on real-time power data, ensuring the stability and efficiency of the flywheel energy storage unit during the charging process, reducing the impact of renewable energy volatility, and improving the reliability and adaptability of the system. In addition, through precise power matching, the energy utilization efficiency is optimized, the service life of the energy storage equipment is effectively extended, while reducing energy waste and promoting better integration and utilization of renewable energy.

[0083] Run data preprocessing includes:

[0084] Data cleaning: The collected operating status data of the flywheel energy storage unit is cleaned to remove noise, outliers and missing data, which is expressed as:

[0085]

[0086] Among them, D clean is the cleaned data set, d i is the i-th piece of data collected, is the mean of the data set, σ is the standard deviation;

[0087] Standardization: The running status data after cleaning is standardized to eliminate the dimensional differences between different parameters, expressed as:

[0088]

[0089] Among them, D norm is the standardized data set, D clean is the cleaned data set, min(D clean ) and max(D clean ) are the minimum and maximum values ​​in the data set respectively;

[0090] Through the above content, the quality and reliability of the data can be significantly improved, noise and outliers can be effectively removed, and the impact of erroneous data on subsequent analysis and decision-making can be avoided. The standardization process eliminates the dimensional differences between different parameters, making the data have consistent scale and comparability.

[0091] The Convolutional Neural Network (CNN) model includes:

[0092] Input data processing: Perform time series conversion on the pre-processed operating status data (including speed, temperature, vibration, and material fatigue parameters) and convert the operating status data x t Convert to time series data X = {x1, x2, ..., x N}, where x1,x2,...,x N are the flywheel state features collected at time 1, 2, ..., N, where N is the size of the time window;

[0093] Convolutional layer design: By using two convolution kernels K1 and K2, with sizes of 1×3, 1×5, 1×7 and 1×2, 1×4, 1×6 respectively, feature maps Y1 and Y2 at different scales are extracted, expressed as:

[0094] Y1 = Conv(X, K1);

[0095] Y2 = Conv(X, K2);

[0096] Among them, Y1 and Y2 are convolution feature maps obtained by convolution kernels K1 and K2 respectively;

[0097] Feature map merging: The feature maps Y1 and Y2 obtained from different convolution kernels are merged to form the final feature map Y, which is expressed as:

[0098] Y = Merge(Y1,Y2);

[0099] Pooling operation: After the output of the convolutional layer, the final feature map is downsampled by using a 2×2 maximum pooling window to reduce the dimension of the data and extract the most important features, which is expressed as:

[0100] Y pool =MaxPooling(Y);

[0101] Among them, Y pool It is the feature map after the pooling operation;

[0102] Fully connected layer and fatigue state prediction: The feature map obtained after convolution and pooling is input into the fully connected layer to predict the material fatigue state, which is expressed as:

[0103] Z=W·Y pool +b;

[0104] Z act =ReLU(Z);

[0105]

[0106] Among them, Z is the output of the fully connected layer, W is the weight matrix, b is the bias term, and Z act is the result after activation, ReLU is the activation function, is the fatigue prediction result, Sigmoid is the activation function;

[0107] Fatigue trend prediction: Based on the results of the material fatigue state prediction, a recurrent neural network (RNN) is used to predict the future fatigue state. It is expressed as:

[0108]

[0109] Among them, h t is the hidden state of the current time step t, which represents the network’s memory of the predicted value at the previous moment, W h is the weight matrix from hidden state to hidden state, W y is the weight matrix for predicting the output value to the hidden state, is the predicted output of the previous moment, b is the bias term, and f is the tanh activation function;

[0110]

[0111] in, is the prediction output of the current time step t, indicating the prediction of fatigue level, W o is the weight matrix from hidden state to output, b ois the bias term of the output layer;

[0112] Through the above content, the fatigue state and trend of the flywheel energy storage unit can be effectively captured, especially in the long-term operation process, the prediction accuracy can be further improved through adaptive adjustment. At the same time, the time series processing capability of the recurrent neural network (RNN) can accurately track and predict future fatigue changes, with strong timing dependence and continuous data optimization capabilities, thus providing effective protection for the safety and reliability of the flywheel energy storage system.

[0113] Running an optimized strategy generation includes:

[0114] Calculation of damage degree of flywheel material: According to the fatigue prediction results of flywheel material, the fatigue cumulative damage model is used to calculate the damage degree D(t) of flywheel material, which is expressed as:

[0115]

[0116] Where D(t) is the accumulated fatigue damage at time t, is the fatigue prediction result of the flywheel material at time t, N i is the fatigue life of the material under the ith stress amplitude, n i is the stress amplitude experienced by the flywheel in the ith cycle, and n is the total number of stress amplitudes;

[0117] Fatigue risk assessment: Based on the calculated damage degree D(t) of the flywheel material, the fatigue risk R(t) is assessed and expressed as:

[0118] R(t)=1-e -αD(t) ;

[0119] Where R(t) is the fatigue risk at time t, and α is the constant of the effect of fatigue damage on life;

[0120] Operation optimization strategy generation: When the fatigue risk R(t) reaches or exceeds the preset fatigue threshold R th When , the operation optimization strategy is started, including adjusting the flywheel speed ω(t) and load distribution P(t), which can be expressed as:

[0121] ω(t)=ω0×(1-βR(t));

[0122] P(t) = P0 × (1-γR(t));

[0123] Where ω(t) is the adjusted flywheel speed, ω0 is the current flywheel speed, P(t) is the adjusted load distribution, P0 is the current load distribution, β and γ are constants adjusted according to the flywheel operating characteristics and fatigue risk;

[0124] Fatigue threshold R thThe setting method specifically includes:

[0125] Collect historical fatigue damage data: Collect the historical flywheel fatigue damage value D of the flywheel energy storage unit historical (i);

[0126] Calculate the average value of historical fatigue damage: Calculate the average value D of historical fatigue damage data avg-historical , as the fatigue damage benchmark of the flywheel under normal operation, is expressed as:

[0127]

[0128] Where N is the number of samples of historical data (data from the past year), D historical (i) is the fatigue damage value of the flywheel material during the i-th operation in history;

[0129] Set a preset threshold coefficient: select an appropriate coefficient w1 according to the design requirements and experience of the flywheel energy storage unit. The coefficient ranges from 0.7 to 0.9, which means that the optimization strategy is triggered when the flywheel material damage reaches 70% to 90% of the historical maximum value;

[0130] Calculate the preset threshold: Use the average value D of historical damage data avg-historical The preset fatigue threshold R of the flywheel fatigue damage is calculated by using the threshold coefficient w1 th , expressed as:

[0131] R th =w1×D avg-historical ;

[0132] Through the above content, the fatigue damage degree and risk of the flywheel are evaluated in real time, and an optimization strategy is automatically generated when the fatigue risk reaches the preset threshold. By adjusting the speed and load distribution of the flywheel, the flywheel life is effectively extended, the operating efficiency is improved, and the risk of system failure is reduced. It not only has intelligent and automated characteristics, but also can achieve energy saving and preventive maintenance, significantly improving the stability and reliability of the flywheel energy storage system.

[0133] The energy scheduling module includes:

[0134] Energy allocation strategy generation: Based on the generated operation optimization strategy (including flywheel speed adjustment and load distribution optimization), combined with the real-time operation status of the flywheel energy storage unit, power grid, and backup energy storage unit, the current power demand and the charging and discharging status of the energy storage unit are dynamically evaluated to generate the optimal energy allocation strategy;

[0135] Energy flow regulation: According to the generated energy distribution strategy, the energy flow between the flywheel energy storage unit, the power grid and the backup energy storage unit is adjusted in real time;

[0136] Load forecasting and historical data optimization: Combine historical load data with real-time input power to predict future load demand and optimize energy scheduling strategies to ensure efficient operation of the system;

[0137] Through the above content, the optimal energy allocation strategy can be generated to ensure the efficient coordination of different electricity sources. In addition, by adjusting the energy flow in real time and optimizing and predicting based on historical load data, it is possible to adapt to changes in power load in advance and further improve the stability and response speed of the system. This optimization strategy effectively reduces energy waste, improves the overall energy efficiency of the system, and ensures the stability and sustainability of energy supply.

[0138] Energy allocation strategy generation includes:

[0139] Power demand assessment: Combine the real-time operating status of the flywheel energy storage unit, the power grid and the backup energy storage unit to dynamically assess the current power demand and calculate the required power P demand (t), expressed as:

[0140] P demand (t) = P load (t)+P charge (t)-P discharge (t);

[0141] Among them, P load (t) is the load power demand at time t, P charge (t) is the charging power of the flywheel energy storage unit, P discharge (t) is the discharge power of the flywheel energy storage unit;

[0142] Optimal energy allocation strategy generation: Based on the dynamically evaluated power demand and the state of the energy storage unit, an optimization algorithm is used to generate the optimal energy allocation strategy S optimal (t), including the energy flow between the flywheel energy storage unit, the grid and the backup energy storage unit, ensures that energy consumption is optimized and system efficiency is improved while meeting load demand, expressed as:

[0143]

[0144] Among them, P i (t) represents the power output of the i-th energy unit, P demand (t) is the total power demand of the system, m is the number of options or choices considered in the optimization process;

[0145] Generate strategy output: The final output energy allocation strategy S optimal (t) It is used to adjust the energy flow between the flywheel energy storage unit, the power grid and the backup energy storage unit to ensure that the system can operate efficiently, maximize energy utilization efficiency, and improve the reliability and stability of the system;

[0146] Through the above content, it is possible to realize a dynamic evaluation of the current power demand and the charging and discharging status of the energy storage unit, and generate the optimal energy allocation strategy based on this, effectively respond to power demand fluctuations and energy storage status changes, ensure the real-time and accuracy of energy scheduling, maximize the stability and efficiency of the system, reduce energy waste, and optimize the use of energy storage resources. In addition, the strategy can adjust the strategy according to the needs in different situations, has strong adaptability and flexibility, and provides an intelligent scheduling solution for energy storage systems under various operating environments.

[0147] Energy flow regulation includes:

[0148] Real-time power output calculation: According to the generated energy allocation strategy, the power output value of the flywheel energy storage unit is calculated, which is expressed as:

[0149] P flywheel (t) = P opt (t)-P grid (t)-P backup (t);

[0150] Among them, P flywheel (t) is the power output of the flywheel energy storage unit at time t, P opt (t) is the optimized power generated according to the energy allocation strategy, P grid (t) is the power flowing into or out of the grid, P backup (t) is the power inflow or outflow of the backup energy storage unit;

[0151] Energy flow adjustment condition judgment: After calculating the power output of the flywheel energy storage unit, it is judged that the power output of the flywheel energy storage unit does not exceed its maximum and minimum power range. If the power output of the flywheel energy storage unit exceeds its rated power range, the power inflow and outflow of the power grid and the backup energy storage unit are adjusted to maintain power balance, which is expressed as:

[0152] P opt (t)≤P flywheel,,max (t) and P flywheel (t)≥P flywheel,,min (t);

[0153] Among them, P flywheel,,max (t) is the maximum power output of the flywheel energy storage unit, P flywheel,,min (t) is the minimum power output of the flywheel energy storage unit;

[0154] Through the above content, efficient energy utilization of the system is ensured. By optimizing the energy distribution between units according to the real-time power demand and the charging and discharging status of the energy storage unit, not only the response speed and stability of the system are improved, but also the waste of electricity can be effectively avoided and the service life of the energy storage unit can be extended. This adjustment method has dynamic adaptability and can automatically optimize and adjust according to different operating conditions and load requirements, thereby ensuring the balance and efficient operation of the entire system.

[0155] Load forecasting and historical data optimization include:

[0156] Historical load data collection: Collect historical load data from the power grid and flywheel energy storage unit;

[0157] Load demand forecasting: Based on the collected historical load data, the ARIMA model is used to perform time series analysis to predict future load demand, which is expressed as:

[0158]

[0159] Among them, L(t+k) is the load demand at the predicted time t+k, which represents the predicted value at the future time, μ is a constant term, which represents the bias of the model, is the autoregressive coefficient, which indicates the impact of historical load data L(t) on current load forecast. L(t+k-1), L(t+k-2),…, L(t+kp) are historical load data. t is the model error or white noise;

[0160] Optimize energy dispatch strategy: Based on the load demand forecast results, combined with real-time power input and energy storage status, the optimization algorithm is used to optimize the energy flow and generate the optimal dispatch strategy, which is expressed as:

[0161]

[0162] Where P(t) is the real-time power input at time t, L(t+k) is the predicted load demand, w2 is the adjustment factor used to balance the load demand and energy loss, Energy Loss is the energy loss caused by the energy storage unit, conversion process or other factors in the power grid, and T is the total number of time periods;

[0163] Energy dispatch strategy implementation: According to the optimized energy dispatch strategy, the energy flow between the flywheel energy storage unit, the power grid and the backup energy storage unit is adjusted;

[0164] Through the above content, the changing trend of load demand can be accurately reflected. Secondly, the ARIMA model is used to predict load demand, so that the system can accurately predict future load demand based on historical load data and the regularity of time series, thereby avoiding sudden load fluctuations or power shortages. By optimizing the energy scheduling strategy, the energy distribution between the flywheel energy storage unit, the power grid and the backup energy storage unit can be intelligently adjusted to ensure that the system can achieve the best match between load and energy at all times. Real-time prediction and optimization of energy scheduling strategies can improve the overall operating efficiency of the system, reduce energy waste, and reduce energy losses. At the same time, it ensures the reliability and economy of power supply under various operating conditions, thereby effectively improving the stability and sustainability of the entire system.

[0165] The present invention covers any substitution, modification, equivalent method and scheme made on the essence and scope of the present invention. In order to make the public have a thorough understanding of the present invention, specific details are described in detail in the following preferred embodiments of the present invention, but those skilled in the art can fully understand the present invention without the description of these details. In addition, in order to avoid unnecessary confusion about the essence of the present invention, well-known methods, processes, procedures, components and circuits are not described in detail.

[0166] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A flywheel energy storage management system for renewable energy integration, characterized in that: It includes condition monitoring module, energy access module, material fatigue prediction and optimization module and energy scheduling module, among which; The state monitoring module collects the operating state of the flywheel energy storage unit in real time, including speed, temperature, vibration and material fatigue parameters; The energy access module receives input power from renewable energy sources and dynamically adjusts input power from different sources using a real-time power matching algorithm; The material fatigue prediction and optimization module uses a convolutional neural network model to predict the fatigue degree of the flywheel material based on the real-time collected operating status of the flywheel energy storage unit, and generates an operation optimization strategy when fatigue risk is predicted in combination with the fatigue cumulative damage model, which specifically includes: Operation data preprocessing: preprocess the collected operation status of the flywheel energy storage unit, including data cleaning and standardization; Material fatigue prediction: Based on the pre-processed operating status, the convolutional neural network model is used to evaluate the fatigue status of the flywheel material in real time and predict the future fatigue trend; Operation optimization strategy generation: Based on the results of material fatigue prediction and combined with the fatigue cumulative damage model, a dynamic operation optimization strategy is generated, including flywheel speed adjustment and load distribution optimization; The energy scheduling module performs dynamic energy distribution among the flywheel energy storage unit, the power grid and the backup energy storage unit based on the generated operation optimization strategy and the operation status of the flywheel energy storage unit, and formulates the optimal energy flow strategy in combination with historical load data and real-time input power.

2. The flywheel energy storage management system for renewable energy integration according to claim 1, characterized in that: The status monitoring module comprises: Speed ​​collection: The speed of the flywheel is measured in real time through the speed sensor; Temperature collection: Use temperature sensors to monitor temperature changes in the flywheel energy storage unit in real time, including motor temperature, bearing temperature, and flywheel internal temperature; Vibration collection: Real-time monitoring of the vibration state of the flywheel energy storage unit through vibration sensors; Material fatigue parameter collection: The strain and fatigue parameters of the flywheel energy storage unit material are monitored in real time through strain sensors.

3. The flywheel energy storage management system for renewable energy integration according to claim 1, characterized in that: The energy access module comprises: Input power reception: receiving power input from solar energy and wind energy through the power access unit; Real-time power matching: Based on the real-time data of power input, a real-time power matching algorithm is used for dynamic adjustment.

4. The flywheel energy storage management system for renewable energy integration according to claim 1, characterized in that: The operation data preprocessing includes: Data cleaning: Clean the collected operating status data of the flywheel energy storage unit to remove noise, outliers and missing data; Standardization: The running status data after cleaning is standardized to eliminate the dimensional differences between different parameters.

5. The flywheel energy storage management system for renewable energy integration according to claim 4, characterized in that: The convolutional neural network model includes: Input data processing: Perform time series conversion on the preprocessed running status data and convert the running status data x t Convert to time series data X = {x1, x2, ..., x N }, where x1,x2,...,x N are the flywheel state features collected at time 1, 2, ..., N, where N is the size of the time window; Convolutional layer design: By using two convolution kernels K1 and K2, with sizes of 1×3, 1×5, 1×7 and 1×2, 1×4, 1×6 respectively, feature maps Y1 and Y2 at different scales are extracted; Feature map merging: Merge the feature maps Y1 and Y2 obtained from different convolution kernels to form the final feature map Y; Pooling operation: After the output of the convolutional layer, the final feature map is downsampled by using a 2×2 maximum pooling window; Fully connected layer and fatigue state prediction: The feature map obtained after convolution and pooling is input into the fully connected layer to predict the material fatigue state; Fatigue trend prediction: Based on the results of material fatigue state prediction, a recurrent neural network is used to predict the future fatigue state.

6. The flywheel energy storage management system for renewable energy integration according to claim 5, characterized in that: The operation optimization strategy generation includes: Calculation of damage degree of flywheel material: Based on the fatigue prediction results of flywheel material, the fatigue cumulative damage model is used to calculate the damage degree D(t) of flywheel material; Fatigue risk assessment: Based on the calculated damage degree D(t) of the flywheel material, the fatigue risk R(t) is assessed; Operation optimization strategy generation: When the fatigue risk R(t) reaches or exceeds the preset fatigue threshold R th When , the operation optimization strategy is started, including adjusting the flywheel speed ω(t) and load distribution P(t).

7. The flywheel energy storage management system for renewable energy integration according to claim 1, characterized in that: The energy scheduling module includes: Energy allocation strategy generation: Based on the generated operation optimization strategy, combined with the real-time operation status of the flywheel energy storage unit, the power grid, and the backup energy storage unit, the current power demand and the charging and discharging status of the energy storage unit are dynamically evaluated to generate the optimal energy allocation strategy; Energy flow regulation: According to the generated energy distribution strategy, the energy flow between the flywheel energy storage unit, the power grid and the backup energy storage unit is adjusted in real time; Load forecasting and historical data optimization: Combine historical load data with real-time input power to predict future load demand and optimize energy scheduling strategies.

8. The flywheel energy storage management system for renewable energy integration according to claim 7, characterized in that: The energy allocation strategy generation includes: Power demand assessment: Combine the real-time operating status of the flywheel energy storage unit, the power grid and the backup energy storage unit to dynamically assess the current power demand and calculate the required power P demand (t); Optimal energy allocation strategy generation: Based on the dynamically evaluated power demand and the state of the energy storage unit, an optimization algorithm is used to generate the optimal energy allocation strategy S optimal (t), including the energy flow between the flywheel energy storage unit, the grid and the backup energy storage unit; Generate strategy output: The final output energy allocation strategy S optimal (t) is used to adjust the energy flow between the flywheel energy storage unit, the power grid and the backup energy storage unit.

9. The flywheel energy storage management system for renewable energy integration according to claim 8, characterized in that: The energy flow regulation includes: Real-time power output calculation: According to the generated energy allocation strategy, the power output value of the flywheel energy storage unit is calculated, which is expressed as: P flywheel (t)=P opt (t)-P grid (t)-P backup (t); Among them, P flywheel (t) is the power output of the flywheel energy storage unit at time t, P opt (t) is the optimized power generated according to the energy allocation strategy, P grid (t) is the power flowing into or out of the grid, P backup (t) is the power inflow or outflow of the backup energy storage unit; Energy flow adjustment condition judgment: After calculating the power output of the flywheel energy storage unit, determine that the power output of the flywheel energy storage unit does not exceed its maximum and minimum power range. If the power output of the flywheel energy storage unit exceeds its rated power range, maintain power balance by adjusting the power inflow and outflow of the power grid and the backup energy storage unit.

10. The flywheel energy storage management system for renewable energy integration according to claim 9, characterized in that: The load forecasting and historical data optimization include: Historical load data collection: Collect historical load data from the power grid and flywheel energy storage unit; Load demand forecasting: Based on the collected historical load data, the ARIMA model is used to perform time series analysis to predict future load demand; Optimize energy dispatch strategy: Based on the load demand forecast results, combined with real-time power input and energy storage status, use optimization algorithms to optimize energy flow and generate the best dispatch strategy; Energy dispatch strategy implementation: According to the optimized energy dispatch strategy, the energy flow between the flywheel energy storage unit, the power grid and the backup energy storage unit is adjusted.

Images