Rapid switching decision management method and system for emergency power supply of thermal power plant

Abstract

The invention relates to a rapid switching decision management method and system for a thermal power plant emergency power supply. The method comprises the following steps: collecting and preprocessing emergency power supply system operation data, equipment vibration data, thermal power plant operation data and external environment data; according to a preset multi-class situation, setting a weight for the multi-source data at each moment and calculating a situation fusion degree; constructing a fault prediction model, and outputting a fault prediction result by taking the preprocessed data and the situation fusion degree as input; setting a plurality of candidate emergency power supply switching schemes such as standby commercial power and a diesel engine based on a prediction result, calculating a comprehensive score of each scheme, and selecting an optimal scheme; and power generation data is monitored in real time, and switching is automatically executed when a switching criterion threshold value corresponding to the scheme is reached, so that rapid, safe and stable switching of the emergency power supply is realized.

Description

TECHNICAL FIELD

[0001] The application relates to a power plant emergency power supply quick switching decision management method and system, and belongs to the technical field of power management. BACKGROUND

[0002] With the continuous improvement of the power industry's requirements for power supply reliability, as a key energy infrastructure, the quick and safe switching of the emergency power supply system of a thermal power plant has become a core link to ensure the stable operation of the unit. In the event of a sudden failure or interruption of the main power supply, the emergency power supply needs to complete the switching from the main power supply to the standby power supply (such as the city power supply or a diesel generator set) in a very short time (usually seconds) to avoid damage to the equipment or expansion of the accident caused by power failure of the boiler feed water pump, the turbine lubricating oil pump and other primary loads. However, the traditional emergency power supply switching system mainly relies on single sensor data or fixed threshold triggering, lacks comprehensive analysis capability of multi-source heterogeneous data, and is difficult to realize real-time sensing of the dynamic changes of the device state, the power grid stability and the environmental factors. In addition, the emergency power supply system of the thermal power plant involves multiple subsystems such as electrical, thermal and mechanical systems, and the data types are complex (such as voltage, current, vibration and temperature), and the requirements of different load levels (primary / secondary / tertiary) for switching time, cost and stability are significantly different, which further increases the complexity of the switching decision. Therefore, developing an intelligent management system capable of integrating multi-source data, dynamically evaluating the system situation and optimizing the switching strategy has become an urgent need to improve the emergency power supply capacity of the thermal power plant.

[0003] The traditional emergency power supply switching technology mainly has the following limitations: firstly, the data acquisition and processing capability is insufficient, most systems only monitor the main power supply state through a single sensor (such as a voltage relay), lack real-time acquisition of key parameters such as device vibration, temperature and power grid frequency, and lead to delayed fault warning or misjudgment; secondly, the decision mechanism is rigid, the traditional scheme mainly adopts fixed threshold (such as triggering switching when the voltage is lower than 85% of the rated value), does not consider the real-time risk of the system, the difference of load levels and the switching cost, and is easy to cause 'excessive switching' or 'delayed switching'; thirdly, it lacks closed-loop optimization capability. The traditional system does not perform recording and analysis on the process data (such as voltage fluctuation and switching time) after switching execution, and cannot feedback to the decision layer to adjust the threshold or strategy, leading to repeated occurrence of similar faults. SUMMARY

[0004] In order to solve the problems existing in the prior art, the application provides a thermal power plant emergency power supply quick switching decision management method and system.

[0005] The technical scheme of the application is as follows: On the one hand, the application provides a thermal power plant emergency power supply quick switching decision management method, which comprises the following steps: Collect operation data of emergency power supply system of thermal power plant, vibration data of emergency power supply system equipment, operation data of thermal power plant and external environment data, and perform preprocessing; Pre-set multiple types of data situation, set weight and situation of each type of data at each time after preprocessing, and calculate different situation fusion degrees of all data at different times based on the weight and situation of each type of data at each time after preprocessing; Construct a fault prediction model, and use the preprocessed data and different situation fusion degrees of all data at different times as inputs of the fault prediction model, and use the fault prediction model to output fault prediction results; Set multiple candidate emergency power supply switching schemes, calculate comprehensive scores of different candidate emergency power supply switching schemes based on the fault prediction results, and select an emergency power supply switching scheme with the highest comprehensive score; Real-time monitor power generation data of the thermal power plant, calculate a switching criterion threshold corresponding to the selected emergency power supply switching scheme, and execute the emergency power supply switching scheme when the power generation data of the thermal power plant reaches the threshold.

[0006] Preferably, the operation data of the emergency power supply system of the thermal power plant, the vibration data of the emergency power supply system equipment, the operation data of the thermal power plant and the external environment data are preliminarily preprocessed by using a sliding average filtering algorithm and a wavelet threshold denoising algorithm, and then the missing values are completed, the abnormal values are removed and the normalization is performed to complete the final preprocessing.

[0007] Preferably, the weight calculation formula of each type of data at each time after preprocessing is:

[0008] Wherein: represents the weight of the i-th type of data at the j-th time; represents the sensitivity coefficient of the i-th type of data at the j-th time; represents the overall situation risk coefficient of the data at the j-th time; represents the importance coefficient of the i-th type of data.

[0009] Preferably, the situation setting step is: Pre-set multiple situations, set a threshold interval for each situation for each type of data, and set the situation of each type of data at different times by the interval of the type of data at different times; The calculation formula of the different situation fusion degrees of all data at different times is:

[0010] Wherein: ​​​express All data at any given moment are relevant to the situation. The degree of integration; Indicates the total number of data types; express Time of the first Weights of class data; express Time of the first Type of data on the situation Support level, specifically 1 and Time of the first Values ​​and Status of Class Data The absolute value of the difference in the deviation rates of the midpoint values ​​within the threshold interval, if Time of the first Class data is not in a state of affairs Threshold range, then Time of the first Type of data on the situation The support rate is 0; Indicates the total number of situations.

[0011] Preferably, the fault prediction model is constructed based on a CNN-LSTM hybrid model.

[0012] Preferably, the fault prediction results include fault type and fault time, and the estimated switching time, estimated switching cost and power supply stability of different candidate emergency power supply switching schemes are set according to different fault types and fault times. The formula for calculating the comprehensive score of different candidate emergency power switching schemes is as follows:

[0013] in: Indicates the current predicted fault result Time of the first The overall score of each candidate emergency power switching scheme; Indicates the switching time weight; Indicates the switching cost weight; This indicates a switch to stability weights; Indicates the current predicted fault result Time of the first Estimated switching time for each candidate emergency power switching scheme; Indicates the switching time base value; Indicates the result of the previous fault prediction Time of the first Estimated switching costs for each candidate emergency power switching solution; This represents the baseline value for switching costs; Indicates the first A power supply stability coefficient of a candidate emergency power switching scheme; represents a risk compensation coefficient.

[0014] Preferably, the switching criterion threshold corresponding to the calculated emergency power switching scheme comprises a voltage switching criterion threshold and a frequency switching criterion threshold. The calculation formula of the voltage switching criterion threshold is:

[0015] Wherein: represents the voltage switching criterion threshold at the moment; represents the voltage switching criterion threshold reference value; represents the first adjustment coefficient; represents the actual load of the thermal power plant unit at the moment; represents the maximum load of the thermal power plant unit; represents the number of equivalent power equipment of the thermal power plant at the moment; represents the maximum number of equivalent power equipment of the thermal power plant; The calculation formula of the frequency switching criterion threshold is:

[0016] Wherein: represents the frequency switching criterion threshold at the moment; represents the second adjustment coefficient; represents the power demand of the load end at the moment; represents the maximum power demand of the load end.

[0017] In another aspect, the present application also provides a thermal power plant emergency power supply rapid switching decision management system, comprising a data acquisition module, a situation fusion module, a fault prediction module, an emergency power supply switching scheme management module and an emergency power supply switching module; The data acquisition module is used to acquire the operation data of the emergency power supply system of the thermal power plant, the vibration data of the emergency power supply system equipment, the operation data of the thermal power plant and the external environment data, and to preprocess them; The situation fusion module is used to preset multiple types of data situations, set the weight and situation of each type of data at each moment after preprocessing, and calculate the different situation fusion degrees of all data at different moments based on the weight and situation of each type of data at each moment after preprocessing. The fault prediction module is used for constructing a fault prediction model, and fusing the preprocessed various types of data and different situation degrees of all data at different moments as inputs of the fault prediction model, and the fault prediction model outputs a fault prediction result; The emergency power switching scheme management module is used for setting a plurality of candidate emergency power switching schemes, calculating comprehensive scores of different candidate emergency power switching schemes based on the fault prediction result, and selecting an emergency power switching scheme with the highest comprehensive score; The emergency power switching module is used for monitoring power generation data of the thermal power plant in real time, calculating a switching criterion threshold value corresponding to the selected emergency power switching scheme, and executing the emergency power switching scheme when the power generation data of the thermal power plant reaches the threshold value.

[0018] In still another aspect, the present application further provides an electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the method according to the present application when executing the program.

[0019] In still another aspect, the present application further provides a computer readable storage medium having a computer program stored thereon, wherein the program is executable on a processor to implement the method according to the present application.

[0020] The present application has the following beneficial effects: 1. The present application realizes real-time and accurate evaluation of the state of the emergency power supply system through the multi-source information fusion module and the dynamic weight distribution algorithm.

[0021] 2. The present application innovatively combines preprocessing technologies such as moving average filtering and wavelet threshold denoising to clean heterogeneous data such as voltage, current, and vibration with high quality, and dynamically adjusts the weight according to the real-time risk and data sensitivity of the system through the dynamic weight distribution algorithm, so that the situation evaluation result is more in line with the actual working condition, which significantly improves the accuracy of fault warning and the reliability of decision-making, and provides a more scientific basis for emergency power switching.

[0022] 3. The present application uses a CNN-LSTM hybrid model to simultaneously capture local features (such as voltage mutation) and time sequence dependence (such as slow rising trend of temperature) of emergency power supply data, and realizes high-precision prediction by combining historical fault cases, solving the limitation of traditional models that can only process single type of data. At the same time, the self-adaptive execution module dynamically optimizes the voltage / frequency switching criterion threshold value by collecting working condition data such as unit load and power demand in real time, forming a closed-loop optimization mechanism of "prediction-decision-execution-feedback". This design enables the system to adapt to switching requirements under different working conditions, significantly improving the stability and response efficiency of the emergency power supply. BRIEF DESCRIPTION OF DRAWINGS

[0023] Figure 1 The figure is a flowchart of the method of the present application. DETAILED DESCRIPTION

[0024] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by a person of ordinary skill in the art without creative effort belong to the scope of protection of the present application.

[0025] It should be understood that the step numbers used herein are only for the convenience of description, and are not limited to the execution sequence of the steps.

[0026] It should be understood that the terms used in the specification of the present application are only for the purpose of describing specific embodiments and are not intended to limit the present application. As used in the specification and the appended claims of the present application, the singular forms "a", "an" and "the" are intended to include the plural forms unless the context clearly indicates otherwise.

[0027] The terms "comprise" and "include" indicate the presence of described features, integers, steps, operations, elements, and / or components, but do not exclude one or more other features, integers, steps, operations, elements, components, and / or groups thereof.

[0028] The term "and / or" means any combination of one or more of the associated listed items and all possible combinations thereof, and includes these combinations.

[0029] Reference Figure 1 A power plant emergency power supply rapid switching decision management method, comprising the following steps: Collecting operation data of an emergency power supply system of a thermal power plant, vibration data of emergency power supply system equipment, operation data of the thermal power plant, and external environment data, and preprocessing; Pre-setting multi-class data situations, setting weights and situations of each type of data at each time after preprocessing, and calculating different situation fusion degrees of all data at different times based on the weights and situations of each type of data at each time after preprocessing; Building a fault prediction model, taking the preprocessed data of each type and the different situation fusion degrees of all data at different times as inputs of the fault prediction model, and outputting fault prediction results from the fault prediction model; Setting multiple candidate emergency power supply switching schemes, calculating comprehensive scores of different candidate emergency power supply switching schemes based on the fault prediction results, and selecting an emergency power supply switching scheme with the highest comprehensive score; Real-time monitoring of power generation data of the thermal power plant, calculating a switching criterion threshold corresponding to a selected emergency power switching scheme, and executing the emergency power switching scheme when the power generation data of the thermal power plant reaches the threshold.

[0030] In one specific embodiment, the real-time operation data of the emergency power supply system includes voltage data of the main power supply inlet, the standby power supply inlet and the diesel generator output collected by voltage sensors, current data of each power supply loop and the power supply loop of key electrical equipment collected by current sensors, and temperature data of the surfaces of contactors, circuit breakers and other key components in the emergency power supply cabinet and the cylinder body of the diesel generator set collected by temperature sensors. The vibration data of the emergency power supply system equipment includes vibration signals of the housings of rotating equipment such as the bearings of the diesel generator set and the motor of the water pump collected by vibration sensors. The operation data of the thermal power plant includes switch states, protection device action signals obtained by interfacing with the electrical control system, parameters such as boiler water level, steam pressure and turbine speed obtained by interfacing with the thermal control system, and device health state data such as transformer oil dissolved gas content and cable insulation resistance obtained by interfacing with the device state monitoring system. The external environment data includes weather data such as rainfall, wind speed and temperature obtained by a meteorological API interface, and power grid stability data such as power grid frequency, voltage deviation and transmission line load rate obtained by a power grid dispatching system interface.

[0031] In one specific embodiment, the installation positions, collected parameters and frequencies of the various sensors are as follows: voltage sensors are installed at the main power supply inlet, the standby power supply inlet and the diesel generator output of the emergency power supply system to collect voltage data at each position, and the collection frequency is set to 10 Hz. Current sensors are installed at each power supply loop (main power supply loop, standby power supply loop and diesel generator set loop) and the power supply loop of key electrical equipment (feed water pump, induced draft fan, transformer, etc.) to collect current data of each loop and device, and the collection frequency is set to 10 Hz. Temperature sensors are attached to the surfaces of contactors, circuit breakers and other key electrical components in the emergency power supply cabinet, as well as the cylinder body of the diesel generator set and the outer wall of the transformer oil tank to collect temperature data of the components and equipment, and the collection frequency is set to 5 Hz. Vibration sensors are fixed by bolts at vibration sensitive parts of rotating equipment such as the bearing seat of the diesel generator set, the housing of the water pump motor and the rotor bearing seat of the fan to collect vibration signals during equipment operation, and the collection frequency is set to 100 Hz. The real-time operating data and equipment vibration signals collected by the aforementioned sensors are combined with internal system data obtained through the system interface (switch status and protection action signals of the electrical control system, boiler water level and steam pressure of the thermal control system, insulation resistance and gas content in oil of the equipment status monitoring system, all collected at a frequency of 1Hz), weather data obtained through the meteorological API interface (rainfall, wind speed, and temperature, collected once every 15 minutes), and power grid stability data obtained through the power grid dispatching system interface (power grid frequency, voltage deviation, and line load rate, collected once every 5 minutes).

[0032] In some embodiments, the operation data of the emergency power system of the thermal power plant, the vibration data of the emergency power system equipment, the operation data of the thermal power plant, and the external environment data are preprocessed by the moving average filtering algorithm and the wavelet threshold denoising algorithm. Then, missing values ​​are filled, outliers are removed, and normalization is performed to complete the final preprocessing.

[0033] In a specific implementation, for time-series data such as voltage, current, and temperature that change slowly, a moving average filter with a window size of 5-10 sampling points is first used to remove high-frequency random noise. Then, the filtered data is checked for trend consistency (if the difference between two adjacent sampling points exceeds 10% of the previous sampling point, it is judged as abnormal and linear interpolation is used for correction). For high-frequency data such as equipment vibration signals, wavelet threshold denoising (using the db4 wavelet basis and decomposition layer 3) is first used to remove interference noise. Then, power spectral density analysis is used to extract the effective vibration signal in the 20-500Hz frequency band to ensure data quality.

[0034] In some embodiments, the weight calculation formula for each type of data at each preprocessed time step is as follows:

[0035] in: express Time of the first Weights of class data; express Time of the first Sensitivity coefficient for class data; express Overall situation risk coefficient based on real-time data; Indicates the first Importance coefficients for class data.

[0036] In some embodiments, the steps for setting the situation are as follows: Multiple scenarios are pre-set, and a threshold range is set for each scenario for each type of data. The scenario for each type of data is set at different times based on the range in which the data falls at different times. The formula for calculating the degree of fusion of different situations of all data at different times is:

[0037] in: express All data at any given moment are relevant to the situation. The degree of integration; Indicates the total number of data types; express Time of the first Weights of class data; express Time of the first Type of data on the situation Support level, specifically 1 and Time of the first Values ​​and Status of Class Data The absolute value of the difference in the deviation rates of the midpoint values ​​within the threshold interval, if Time of the first Class data is not in a state of affairs Threshold range, then Time of the first Type of data on the situation The support rate is 0; Indicates the total number of situations.

[0038] In one specific embodiment, three states are set: "normal," "minor anomaly," and "severe anomaly." For voltage data, the preset threshold range for the "normal" state is 90%-110% of the rated voltage; the preset threshold range for the "minor anomaly" state is 85%-90% and 110%-115% of the rated voltage; and the preset threshold range for the "severe anomaly" state is below 85% or above 115% of the rated voltage. For current data, the preset threshold range for the "normal" state is 0%-80% of the rated current; the preset threshold range for the "minor anomaly" state is 80%-100% of the rated current; and the preset threshold range for the "severe anomaly" state is... The threshold range is 100% or more of the rated current. For temperature data, the preset threshold range for "normal" condition is 70% to 70% of the maximum allowable temperature of the equipment, the threshold range for "slightly abnormal" condition is 70%-90% of the maximum allowable temperature of the equipment, and the threshold range for "severely abnormal" condition is more than 90% of the maximum allowable temperature of the equipment. For vibration signals, the preset threshold range for "normal" condition is 0%-80% of the factory vibration standard value of the equipment, the threshold range for "slightly abnormal" condition is 80%-100% of the factory vibration standard value of the equipment, and the threshold range for "severely abnormal" condition is more than 100% of the factory vibration standard value of the equipment.

[0039] In one specific embodiment, the overall risk coefficient of the data at different times is the proportion of data with severe abnormal situations.

[0040] In some embodiments, the failure prediction model is constructed based on a CNN-LSTM hybrid model, combines historical failure case data and collected data for inference, and outputs a failure type and a failure prediction occurrence time; In one specific embodiment, the CNN part network structure is set to 3 convolutional layers, 2 pooling layers, and 1 flatten layer. The first convolutional layer uses a 3x3 size convolutional kernel, the number is 64, the step is set to 1, and the activation function uses ReLU, which is used to extract basic local features such as voltage mutation and temperature peak in the data. The second convolutional layer uses a 5x5 size convolutional kernel, the number is 128, the step is set to 1, and the activation function uses ReLU, which is used to further extract complex local features such as specific frequency bands of vibration signals and current fluctuation periods in the data. The third convolutional layer uses a 3x3 size convolutional kernel, the number is 256, the step is set to 1, and the activation function uses ReLU, which is used to fuse the features of the previous two layers and extract more fine local features (such as subtle parameter abnormal features in the early stage of device aging). A max pooling layer is set between adjacent convolutional layers, the pooling kernel size is 2x2, and the step is set to 2, which is used to downsample the feature map output by the convolutional layer, reduce the number of parameters, and preserve the key features. Finally, the two-dimensional feature map output by the pooling layer is converted into a one-dimensional feature vector through the flatten layer, which is used as the input of the LSTM part. In one specific embodiment, the network structure of the LSTM part is set to 2 hidden layers and 1 fully connected layer, the number of neurons in each hidden layer is set to 128, the activation function uses tanh, and the activation functions of the forget gate, input gate, and output gate all use sigmoid. The one-dimensional feature vector output by the CNN is divided into a time sequence feature sequence according to the time step (set to 60, corresponding to the sampling data within 60 seconds), which is input into the first LSTM hidden layer to capture the short-term time sequence dependence of the data (such as the continuous fluctuation trend of voltage within 1 minute); the output of the first LSTM hidden layer is used as the input of the second LSTM hidden layer to capture the long-term time sequence dependence of the data (such as the slow rising trend of temperature within 10 minutes and the periodic change rule of vibration signals); the output of the second LSTM hidden layer is connected to the fully connected layer, the number of neurons in the fully connected layer is set to 64, and the activation function uses ReLU, which is used to integrate the time sequence features extracted by the LSTM.

[0041] In one specific embodiment, the CNN-LSTM hybrid model is trained based on historical failure data of a thermal power plant (at least 3 years, containing 100+ failure cases), uses a "7:3" division of training set / test set, iterates 100 rounds using an Adam optimizer (learning rate 0.001), until the loss function (classification loss + regression loss) converges (loss value <0.05), to ensure the generalization ability of the model.

[0042] In some embodiments, the fault prediction results include fault type and fault time, and the estimated switching time, estimated switching cost and power supply stability of different candidate emergency power supply switching schemes are set according to different fault types and fault times. The formula for calculating the comprehensive score of different candidate emergency power switching schemes is as follows:

[0043] in: Indicates the current predicted fault result Time of the first The overall score of each candidate emergency power switching scheme; Indicates the switching time weight; Indicates the switching cost weight; This indicates a switch to stability weights; Indicates the current predicted fault result Time of the first Estimated switching time for each candidate emergency power switching scheme; Indicates the switching time base value; Indicates the result of the previous fault prediction Time of the first Estimated switching costs for each candidate emergency power switching solution; This represents the baseline value for switching costs; Indicates the first The power supply stability coefficient of each candidate emergency power switching scheme; This represents the risk compensation coefficient.

[0044] In one specific embodiment, two candidate switching schemes, namely backup mains power and diesel generator set, are set. The estimated switching time, estimated switching cost and power supply stability of the two candidate emergency power switching schemes are set according to different fault types and fault times. The table below shows the estimated switching time, estimated switching cost, and power supply stability under certain fault conditions:

[0045] In one specific embodiment, the switching time reference value is set according to the maximum power outage time allowed in the emergency power supply system design specification. Specifically, the switching time reference value for primary loads that are sensitive to power outages (such as boiler feed water pumps and turbine lubricating oil pumps) is set to 5 seconds; the switching time reference value for secondary loads that are more tolerant to power outages (such as auxiliary workshop ventilators) is set to 10 seconds; and the switching time reference value for tertiary loads that have no special requirements for power outages (such as lighting systems) is set to 15 seconds. The switching time weight, switching cost weight, and switching stability weight are set to 0.4, 0.3, and 0.3 respectively, based on the priority of emergency power supply for thermal power plants.

[0046] In some embodiments, the calculated emergency power switching scheme corresponding to the switching criterion threshold includes a voltage switching criterion threshold and a frequency switching criterion threshold; The formula for calculating the voltage switching criterion threshold is as follows:

[0047] in: express Voltage switching criterion threshold at any given time; This represents the threshold reference value for voltage switching criteria; Indicates the first adjustment factor; express The actual load of thermal power plant units at all times; This indicates the maximum load of the thermal power plant unit; express The number of equivalent electrical equipment in a thermal power plant at any given time; This indicates the maximum equivalent number of electrical equipment in a thermal power plant. The formula for calculating the frequency switching criterion threshold is as follows:

[0048] in: express Frequency switching criterion threshold at any given time; Indicates the second adjustment factor; express The electricity demand at the load end at any given time; This indicates the maximum electricity demand at the load end.

[0049] In one specific embodiment, during emergency power switching, real-time change curves of voltage, current, and frequency are recorded at a sampling frequency of 100Hz (the recording duration is from 5 seconds before switching to 10 seconds after switching). After switching is completed, the recorded data is automatically analyzed to calculate the voltage fluctuation amplitude, frequency fluctuation amplitude, and switching time (the time from issuing the switching command to voltage stabilization) during the switching process, and compared with preset switching quality standards (such as voltage fluctuation ≤ ±2%, frequency fluctuation ≤ ±0.5Hz, and switching time ≤ 15 seconds). If the standards are not met, the reasons are analyzed (such as unreasonable threshold adjustment or equipment response delay), and the analysis results are fed back to the decision-making module and the multi-source information fusion module to optimize subsequent scheme decisions and threshold adjustment algorithms.

[0050] In some embodiments, a decision management system for rapid emergency power switching in thermal power plants is proposed, including a data acquisition module, a situation fusion module, a fault prediction module, an emergency power switching scheme management module, and an emergency power switching module. The data acquisition module is used to collect operational data of the emergency power system of the thermal power plant, vibration data of the emergency power system equipment, operational data of the thermal power plant, and external environmental data, and to perform preprocessing. The situation fusion module is used to preset multiple types of data situations, set the weights and situations of various types of data at each time after preprocessing, and calculate the different situation fusion degrees of all data at different times based on the weights and situations of various types of data at each time after preprocessing. The fault prediction module is used to construct a fault prediction model. It takes the preprocessed data and the different states of all data at different times as input to the fault prediction model, and the fault prediction model outputs the fault prediction result. The emergency power switching scheme management module is used to set multiple candidate emergency power switching schemes, calculate the comprehensive score of different candidate emergency power switching schemes based on the fault prediction results, and select the emergency power switching scheme with the highest comprehensive score. The emergency power switching module is used to monitor the power generation data of the thermal power plant in real time, calculate the switching criterion threshold corresponding to the selected emergency power switching scheme, and execute the emergency power switching scheme when the power generation data of the thermal power plant reaches the threshold.

[0051] In some embodiments, an electronic device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the method as described in any embodiment of the present invention.

[0052] In some embodiments, a computer-readable storage medium is provided on which a computer program is stored, which, when executed by a processor, implements the method as described in any embodiment of the present invention.

[0053] In this application embodiment, "at least one" refers to one or more, and "more than one" refers to two or more. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent the existence of A alone, A and B simultaneously, or B alone. A and B can be singular or plural. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. "At least one of the following" and similar expressions refer to any combination of these items, including any combination of singular or plural items. For example, at least one of a, b, and c can represent: a, b, c, a and b, a and c, b and c, or a and b and c, where a, b, and c can be single or multiple.

[0054] Those skilled in the art will recognize that the units and algorithm steps described in the embodiments disclosed herein can be implemented using electronic hardware, computer software, or a combination of electronic hardware and software. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0055] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0056] In the several embodiments provided in this application, any function, if implemented as a software functional unit and sold or used as an independent product, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0057] The above description is merely an embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structural or procedural transformations made based on the content of the present invention's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A method for rapid decision-making and management of emergency power supply switching in thermal power plants, characterized in that, Includes the following steps: Collect operational data of the emergency power supply system of the thermal power plant, equipment vibration data of the emergency power supply system, operational data of the thermal power plant, and external environmental data, and perform preprocessing. Multiple data states are preset, and the weights and states of each type of data at each time point after preprocessing are set. Based on the weights and states of each type of data at each time point after preprocessing, the different states fusion degrees of all data at different times are calculated. A fault prediction model is constructed, and the preprocessed data and the different states of all data at different times are used as inputs to the fault prediction model. The fault prediction model outputs the fault prediction result. Multiple candidate emergency power switching schemes are set up, and the comprehensive score of different candidate emergency power switching schemes is calculated based on the fault prediction results. The emergency power switching scheme with the highest comprehensive score is selected. Real-time monitoring of power generation data from thermal power plants is performed, and the switching criterion threshold corresponding to the selected emergency power switching scheme is calculated. When the power generation data of thermal power plants reaches the threshold, the emergency power switching scheme is executed.

2. The method for rapid switching decision management of emergency power supply in thermal power plants according to claim 1, characterized in that, The operation data of the emergency power system of the thermal power plant, the vibration data of the emergency power system equipment, the operation data of the thermal power plant, and the external environment data are initially preprocessed by the moving average filtering algorithm and the wavelet threshold denoising algorithm. Then, missing values ​​are filled, outliers are removed, and normalization is performed to complete the final preprocessing.

3. The method for rapid switching decision management of emergency power supplies in thermal power plants according to claim 1, characterized in that, The formulas for calculating the weights of various data types at each time step after preprocessing are as follows: in: express Time of the first Weights of class data; express Time of the first Sensitivity coefficient for class data; express Overall situation risk coefficient based on real-time data; Indicates the first Importance coefficients for class data.

4. The method for rapid switching decision management of emergency power supply in thermal power plants according to claim 1, characterized in that, The steps for setting up the situation are as follows: Multiple scenarios are pre-set, and a threshold range is set for each scenario for each type of data. The scenario for each type of data is set at different times based on the range in which the data falls at different times. The formula for calculating the degree of fusion of different situations of all data at different times is: in: express All data at any given moment are relevant to the situation. The degree of integration; Indicates the total number of data types; express Time of the first Weights of class data; express Time of the first Type of data on the situation Support level, specifically 1 and Time of the first Values ​​and Status of Class Data The absolute value of the difference in the deviation rates of the midpoint values ​​within the threshold interval, if Time of the first Class data is not in a state of affairs Threshold range, then Time of the first Type of data on the situation The support rate is 0; Indicates the total number of situations.

5. The method for rapid switching decision management of emergency power supply in thermal power plants according to claim 1, characterized in that, The fault prediction model is built based on a CNN-LSTM hybrid model.

6. The method for rapid switching decision management of emergency power supply in thermal power plants according to claim 1, characterized in that, The fault prediction results include fault type and fault time. Based on different fault types and fault times, the estimated switching time, estimated switching cost, and power supply stability of different candidate emergency power supply switching schemes are set. The formula for calculating the comprehensive score of different candidate emergency power switching schemes is as follows: in: Indicates the current predicted fault result Time of the first The overall score of each candidate emergency power switching scheme; Indicates the switching time weight; Indicates the switching cost weight; This indicates a switch to stability weights; Indicates the current predicted fault result Time of the first Estimated switching time for each candidate emergency power switching scheme; Indicates the switching time base value; Indicates the result of the previous fault prediction Time of the first Estimated switching costs for each candidate emergency power switching solution; This represents the baseline value for switching costs; Indicates the first The power supply stability coefficient of each candidate emergency power switching scheme; This represents the risk compensation coefficient.

7. The method for rapid switching decision management of emergency power supply in thermal power plants according to claim 1, characterized in that, The calculated emergency power switching scheme corresponding to the switching criterion threshold includes voltage switching criterion threshold and frequency switching criterion threshold; The formula for calculating the voltage switching criterion threshold is as follows: in: express Voltage switching criterion threshold at any given time; This represents the threshold reference value for voltage switching criteria; Indicates the first adjustment factor; express The actual load of thermal power plant units at all times; This indicates the maximum load of the thermal power plant unit; express The number of equivalent electrical equipment in a thermal power plant at any given time; This indicates the maximum equivalent number of electrical equipment in a thermal power plant. The formula for calculating the frequency switching criterion threshold is as follows: in: express Frequency switching criterion threshold at any given time; Indicates the second adjustment factor; express The electricity demand at the load end at any given time; This indicates the maximum electricity demand at the load end.

8. A decision-making and management system for rapid switching of emergency power supplies in thermal power plants, characterized in that, It includes a data acquisition module, a situation fusion module, a fault prediction module, an emergency power switching scheme management module, and an emergency power switching module; The data acquisition module is used to collect operational data of the emergency power system of the thermal power plant, vibration data of the emergency power system equipment, operational data of the thermal power plant, and external environmental data, and to perform preprocessing. The situation fusion module is used to preset multiple types of data situations, set the weights and situations of various types of data at each time after preprocessing, and calculate the different situation fusion degrees of all data at different times based on the weights and situations of various types of data at each time after preprocessing. The fault prediction module is used to construct a fault prediction model. It takes the preprocessed data and the different states of all data at different times as input to the fault prediction model, and the fault prediction model outputs the fault prediction result. The emergency power switching scheme management module is used to set multiple candidate emergency power switching schemes, calculate the comprehensive score of different candidate emergency power switching schemes based on the fault prediction results, and select the emergency power switching scheme with the highest comprehensive score. The emergency power switching module is used to monitor the power generation data of the thermal power plant in real time, calculate the switching criterion threshold corresponding to the selected emergency power switching scheme, and execute the emergency power switching scheme when the power generation data of the thermal power plant reaches the threshold.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the method as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the method as described in any one of claims 1 to 7.

Images