A mechanical overspeed monitoring system and method for a hydroelectric generator set

Abstract

A kind of mechanical overspeed monitoring system and method of hydroelectric generating set, including fastening ring, fastening ring is connected with main shaft, overspeed swing and counterweight are installed on fastening ring, overspeed swing is oppositely installed with hydraulic switching valve, the state of overspeed swing and hydraulic switching valve is monitored by high-speed camera;Hydraulic switching valve and high-speed camera are electrically connected with mechanical overspeed monitoring host, and mechanical overspeed monitoring host is electrically connected with vibration acquisition system and rotational speed acquisition system.The present application has the beneficial effects of improving monitoring accuracy, ensuring stable operation of system, improving equipment safety, facilitating fault diagnosis and maintenance and the like, can effectively guarantee the safe, stable and efficient operation of hydroelectric generating set, and provides strong technical support for long-term stable operation of unit.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of state monitoring of hydroelectric generating set, and particularly relates to a mechanical overspeed monitoring system and method for hydroelectric generating set. BACKGROUND

[0002] The hydroelectric generator is generally provided with a mechanical overspeed protection device for the last protection of the unit. The action value of the mechanical overspeed protection device has strict requirements, which aims to effectively act and not to cause the unit to be stopped due to misoperation. At present, the action accuracy of the mechanical overspeed device of the unit only depends on the factory product setting, and the state monitoring of the mechanical overspeed action after the overspeed test of the power station and the operation of the unit is in a blank state.

[0003] In order to timely monitor and early warn the mechanical overspeed, a high-reliability and high-sensitivity mechanical overspeed monitoring system needs to be designed. The technical difficulties that need to be overcome by the system mainly include that the mechanical overspeed monitoring system needs to collect the speed signal of the hydroelectric generating set in real time, and judge whether the mechanical overspeed occurs according to the speed change trend. However, in the high-speed rotating environment, the collection of the speed signal is easily affected by electromagnetic interference and mechanical vibration, resulting in signal distortion or interruption, which affects the accuracy and reliability of the monitoring. SUMMARY

[0004] The technical problem to be solved by the present application is to provide a mechanical overspeed monitoring system and method for hydroelectric generating set, for data collection and analysis during mechanical overspeed, and to improve the product manufacturing and installation reliability and the unit operation safety.

[0005] To solve the above technical problems, the technical scheme adopted by the present application is:

[0006] A mechanical overspeed monitoring system for hydroelectric generating set, comprising a fastening ring connected with a main shaft, an overspeed pendulum and a counterweight mounted on the fastening ring, the overspeed pendulum being oppositely mounted with a hydraulic switching valve, and the state of the overspeed pendulum and the hydraulic switching valve being monitored by a high-speed camera;

[0007] The hydraulic switching valve and the high-speed camera are electrically connected with a mechanical overspeed monitoring host, and the mechanical overspeed monitoring host is electrically connected with a vibration and swing collection system and a speed collection system.

[0008] Preferably, the fastening ring and the main shaft are connected by clamping.

[0009] Preferably, the overspeed pendulum and the counterweight are symmetrically mounted on the fastening ring.

[0010] Preferably, the high-speed camera is fixed by a support, so that the high-speed camera can integrate the hydraulic switching valve and the overspeed pendulum into the camera range.

[0011] Preferably, the swing acquisition system and the rotating speed acquisition system are connected with the mechanical overspeed monitoring host through a data transmission line to ensure that relevant data signals are transmitted.

[0012] Preferably, the mechanical overspeed monitoring host comprises a first signal receiving unit, a second signal receiving unit, a power module, a display screen and a rack, the first signal receiving unit is installed at the lower left side of the installation rack, the second signal receiving unit is installed at the lower middle side of the installation rack, the power module is installed at the lower right side of the installation rack, and the display screen is embedded in the top of the rack.

[0013] A method for operating a mechanical overspeed monitoring system of a hydroelectric generating set, comprising the following steps:

[0014] A high-precision rotating speed sensor is used to acquire the rotating speed signals of the hydroelectric generating set in real time, and a filtering algorithm is used to eliminate the influence of electromagnetic interference and mechanical vibration on the signals to obtain stable rotating speed data.

[0015] The filtered rotating speed data is input into a preset rotating speed change trend analysis model, and if the rotating speed change rate exceeds a preset threshold, it is judged as a mechanical overspeed state.

[0016] When the mechanical overspeed state is judged, a hydraulic switching valve control module is triggered, the switching time and force of the hydraulic switching valve are calculated according to the moment of inertia of the hydroelectric generating set, and the water hammer effect is avoided.

[0017] The hydraulic switching valve control module outputs a control signal to a hydraulic actuator to quickly switch the state of the hydraulic switching valve and realize emergency shutdown.

[0018] During the switching process of the hydraulic switching valve, the pressure change of the hydraulic system is monitored in real time, and if the pressure fluctuation exceeds a preset range, the switching parameters of the hydraulic switching valve are adjusted to ensure the stability of the system.

[0019] The system is provided with a fault self-diagnosis module, which periodically detects the working state of the rotating speed sensor, the hydraulic switching valve control module and the signal processing module, and if a fault is detected, a fault locking function is triggered to prevent false alarms.

[0020] The fault self-diagnosis module stores fault information in the system log and uploads it to the remote monitoring center through the communication module for subsequent analysis and maintenance.

[0021] Preferably, a high-precision rotating speed sensor is used to acquire the rotating speed signals of the hydroelectric generating set in real time, and a filtering algorithm is used to eliminate the influence of electromagnetic interference and mechanical vibration on the signals to obtain stable rotating speed data, including:

[0022] A high-precision rotating speed sensor is used to acquire the rotating speed signals of the hydroelectric generating set in real time to obtain original rotating speed data.

[0023] For the collected original speed data, wavelet transform algorithm is used for denoising processing to eliminate the influence of electromagnetic interference on the speed signal;

[0024] According to the mechanical structure characteristics of the hydro-generator unit, a mechanical vibration mathematical model is established, and the speed signal is filtered through an adaptive filtering algorithm to eliminate the influence of mechanical vibration on the speed signal;

[0025] The denoised and filtered speed signal is subjected to feature extraction to obtain time domain and frequency domain characteristic parameters of the speed signal;

[0026] The extracted speed signal characteristic parameters are input into a support vector machine model, and through model training and optimization, a stable and reliable speed estimation value is obtained;

[0027] The estimated speed value is compared with the preset speed threshold value, and if the speed exceeds the normal range, it is judged that the hydro-generator unit is abnormal, and an alarm signal is triggered;

[0028] According to the speed estimation value and the alarm signal, a running state report of the hydro-generator unit is generated to provide data support for equipment maintenance and fault diagnosis.

[0029] Preferably, the filtered speed data is input into a preset speed change trend analysis model, and if the speed change rate exceeds the preset threshold value, it is judged that the mechanical overspeed state, including:

[0030] Real-time speed data of the mechanical equipment is obtained, the speed data is subjected to filtering processing, and noise interference in the speed data is removed to obtain filtered speed data;

[0031] According to the filtered speed data, the speed change rate is calculated to obtain speed change rate data within a certain time range;

[0032] The speed change rate data is input into a preset speed change trend analysis model, and the speed change rate data is subjected to trend analysis through the model to judge whether the speed change rate exceeds the preset threshold value;

[0033] If the speed change rate exceeds the preset threshold value, it is judged that the mechanical equipment is in an overspeed state, an overspeed alarm is triggered, and relevant personnel are notified for processing;

[0034] If the speed change rate does not exceed the preset threshold value, it is judged that the mechanical equipment is in a normal working state, real-time speed data is continuously obtained, and the next round of speed change trend analysis is performed;

[0035] According to the historical speed change rate data and the overspeed state judgment result, a support vector machine algorithm is used to train and optimize the speed change trend analysis model to improve the accuracy of the model in judging the overspeed state;

[0036] By big data analysis technology, the association rules between the speed change rate and other operating parameters of the mechanical equipment are mined, and the speed change trend analysis model is optimized combined with the association rules, so as to improve the adaptability and robustness of the model.

[0037] Preferably, when the mechanical overspeed state is determined, the hydraulic switching valve control module is triggered, the switching time and force of the hydraulic switching valve are calculated according to the moment of inertia of the hydro-generator unit, and the water hammer effect is avoided, including:

[0038] The speed of the hydro-generator unit is monitored in real time by the speed sensor, and the collected speed data is transmitted to the mechanical overspeed monitoring host;

[0039] After receiving the speed data, the mechanical overspeed monitoring host compares it with the preset mechanical overspeed threshold to determine whether the current is in the mechanical overspeed state;

[0040] If the determination result is the mechanical overspeed state, the mechanical overspeed monitoring host immediately sends a trigger signal to the hydraulic switching valve control module;

[0041] After receiving the trigger signal, the hydraulic switching valve control module obtains the moment of inertia parameter of the hydro-generator unit from the equipment parameter database;

[0042] According to the obtained moment of inertia parameter, the optimal switching time and switching force of the hydraulic switching valve are calculated by using an optimization algorithm, so as to realize rapid and effective mechanical deceleration while avoiding the water hammer effect;

[0043] The calculated hydraulic switching valve switching time and force parameters are transmitted to the hydraulic switching valve execution unit to control the hydraulic switching valve to switch according to the optimal scheme;

[0044] After the hydraulic switching valve is switched, the speed change of the hydro-generator unit is continuously monitored until the speed returns to the normal range, so as to ensure that the mechanical overspeed problem is effectively solved.

[0045] Preferably, the hydraulic switching valve control module outputs a control signal to the hydraulic actuator to quickly switch the state of the hydraulic switching valve and realize emergency shutdown, including:

[0046] The hydraulic switching valve control module receives the emergency shutdown instruction and determines whether the emergency shutdown operation needs to be performed, and if the emergency shutdown operation needs to be performed, the next step is entered;

[0047] Otherwise, the emergency shutdown instruction is continuously monitored;

[0048] The hydraulic switching valve control module determines the hydraulic switching valve state that needs to be switched according to the preset hydraulic switching valve switching control strategy, and generates a corresponding hydraulic switching valve control signal;

[0049] The hydraulic switching valve control module outputs the generated hydraulic switching valve control signal to the hydraulic actuator, and drives the hydraulic actuator to quickly switch the state of the hydraulic switching valve through the control signal;

[0050] After receiving the control signal output by the hydraulic switching valve control module, the hydraulic actuator quickly switches the state of the hydraulic switching valve according to the control signal, realizing the quick switching of the hydraulic switching valve;

[0051] The hydraulic switching valve control module monitors the state feedback signal of the hydraulic actuator in real time, judges whether the hydraulic switching valve has been switched to the target state, and if so, proceeds to the next step;

[0052] Otherwise, continue to output the control signal to drive the hydraulic actuator to switch the state of the hydraulic switching valve;

[0053] After determining that the hydraulic switching valve has been switched to the target state according to the state feedback signal of the hydraulic actuator, the hydraulic switching valve control module stops outputting the control signal and feeds back the switching result of the hydraulic switching valve state to the upper control system;

[0054] After receiving the switching result of the hydraulic switching valve state fed back by the hydraulic switching valve control module, the upper control system judges whether the emergency shutdown operation has been completed, and if so, ends the emergency shutdown process;

[0055] Otherwise, continue to perform other emergency shutdown operations.

[0056] Preferably, during the switching process of the hydraulic switching valve, the pressure change of the hydraulic system is monitored in real time, and if the pressure fluctuation exceeds the preset range, the switching parameters of the hydraulic switching valve are adjusted to ensure system stability, including:

[0057] Real-time pressure data during the switching process of the hydraulic switching valve are obtained, and the pressure data are compared with the preset range to determine whether the pressure fluctuation exceeds the preset range;

[0058] If the pressure fluctuation exceeds the preset range, the adjustment range of the switching parameters of the hydraulic switching valve is determined according to the degree of the pressure fluctuation exceeding the range;

[0059] An association model between the pressure fluctuation and the switching parameters of the hydraulic switching valve is established through a machine learning algorithm, and the optimal switching parameters of the hydraulic switching valve are predicted according to the pressure fluctuation;

[0060] The adjusted switching parameters of the hydraulic switching valve are applied to the switching process of the hydraulic switching valve, and the pressure change is continuously monitored to determine whether the adjusted parameters are effective;

[0061] If the adjusted parameters fail to effectively reduce the pressure fluctuation, the machine learning model is further optimized, and the optimal switching parameters of the hydraulic switching valve are re-predicted;

[0062] The hydraulic switching valve switching parameters are continuously iteratively optimized until the pressure fluctuation is controlled within the preset range, ensuring stable operation of the hydraulic system;

[0063] The optimized hydraulic switching valve switching parameters are saved as preset parameters, which are directly applied in the subsequent hydraulic switching valve switching process, improving the system response speed and stability.

[0064] Preferably, the system is built-in with a fault self-diagnosis module that periodically detects the working state of the speed sensor, hydraulic switching valve control module and signal processing module. If a fault is detected, the fault lock function is triggered to prevent false alarms, including:

[0065] According to the preset time interval, the system built-in fault self-diagnosis module is triggered periodically to detect the state of the speed sensor, hydraulic switching valve control module and signal processing module;

[0066] When performing state detection, the fault self-diagnosis module obtains real-time speed data of the speed sensor and compares it with the preset normal speed range to determine whether the speed sensor has a fault;

[0067] At the same time, the fault self-diagnosis module analyzes the feedback signal of the hydraulic switching valve control module to determine whether the opening and closing state of the hydraulic switching valve is consistent with the control command, and determines whether the hydraulic switching valve control module has a fault;

[0068] The fault self-diagnosis module also monitors the input and output data of the signal processing module in real time, analyzes the continuity and integrity of the data, and determines whether the signal processing module is working normally;

[0069] If the fault self-diagnosis module detects that any module has a fault, the fault lock function is immediately triggered to stop the work of the corresponding module by cutting off the power supply or blocking the signal transmission;

[0070] At the same time of triggering the fault lock function, the fault self-diagnosis module generates a fault report, records the time and location of the fault occurrence and the fault type, etc. information, and sends the report to the system administrator for subsequent maintenance;

[0071] The fault lock function continues to take effect until the system administrator confirms that the fault has been eliminated and manually releases the lock state, and the system can resume normal work, thereby effectively preventing false alarms caused by the continued operation of the faulty module.

[0072] Preferably, the fault self-diagnosis module stores the fault information to the system log and uploads it to the remote monitoring center through the communication module for subsequent analysis and maintenance, including:

[0073] According to the fault information obtained by the fault self-diagnosis module, the fault information is written into the system log to obtain a system log file containing the fault information;

[0074] A communication connection is established between the communication module and the remote monitoring center, and if the communication connection is successfully established, the system log file is uploaded to the remote monitoring center;

[0075] After the remote monitoring center receives the uploaded system log file, the system log file is analyzed using natural language processing technology, and key fault information is extracted;

[0076] According to the extracted fault information, semantic analysis is performed through knowledge graph technology to determine the fault type and fault cause, and a fault diagnosis result is obtained;

[0077] The fault diagnosis result is matched with the pre-established fault solution knowledge base to obtain a corresponding fault solution;

[0078] According to the obtained fault solution, a fault maintenance work order is automatically generated, and the maintenance work order is distributed to the relevant maintenance personnel, and the maintenance personnel are notified to perform fault maintenance;

[0079] After completing the fault maintenance, the maintenance personnel feed back the maintenance result to the remote monitoring center, and the remote monitoring center updates the fault solution knowledge base to optimize the fault diagnosis and solution process.

[0080] The present application can achieve the following beneficial effects:

[0081] 1. The fastening ring and the main shaft are connected by clamping, which can ensure the stable installation of the fastening ring on the main shaft, avoid loosening of the fastening ring due to vibration and other factors during operation of the hydroelectric generator set, and ensure the stability of the installation foundation of the overspeed pendulum and the counterweight, thereby providing reliable hardware support for subsequent overspeed monitoring work.

[0082] 2. The overspeed pendulum and the counterweight are symmetrically installed on the fastening ring, which can make the overall stress of the fastening ring more uniform, avoid additional centrifugal force or other additional forces due to asymmetric installation, ensure the balance and stability of the overspeed monitoring system during operation, and improve the accuracy and reliability of monitoring.

[0083] 3. The high-speed camera is fixed by a bracket, which can integrate the hydraulic switching valve and the overspeed pendulum into the camera range, facilitate unified real-time monitoring of the state of these two key components, help to grasp the overall system operation state, and also facilitate fault troubleshooting and analysis of possible problems in the later stage.

[0084] 4. The vibration and rotation speed acquisition systems are connected to the mechanical overspeed monitoring host via data transmission lines to ensure that relevant data signals can be accurately and stably transmitted to the mechanical overspeed monitoring host, reducing interference and loss during signal transmission and providing data support for subsequent data analysis and processing.

[0085] 5. The mechanical overspeed monitoring host includes a first signal receiving unit, a second signal receiving unit, a power supply module, a display screen, and a frame. The internal units are rationally arranged, with the signal receiving unit and power supply module installed in different positions at the bottom of the frame, and the display screen embedded in the top of the frame. This layout not only facilitates the maintenance and management of each component, but also promotes the orderly reception, processing, and display of signals, while avoiding mutual interference between components, thus improving the overall performance and working efficiency of the host.

[0086] 6. High-precision speed sensors are used to acquire the speed signals of the hydro-generator unit in real time. Combined with wavelet transform algorithms and an adaptive filtering algorithm based on the mechanical structure characteristics, the influence of electromagnetic interference and mechanical vibration on the speed signals can be effectively eliminated, obtaining stable and reliable speed data. This helps to monitor the unit speed more accurately, avoid misjudgments caused by interference signals, and improve monitoring accuracy.

[0087] 7. By extracting features from the speed signal and inputting the extracted feature parameters into a support vector machine model for training and optimization, a stable speed estimate can be obtained, further improving the accuracy and reliability of speed monitoring. Simultaneously, based on the comparison between the speed estimate and a preset threshold, abnormal conditions of the hydro-generator unit can be detected promptly and alarm signals can be triggered, providing effective data support for equipment maintenance and fault diagnosis, and helping to prevent equipment failures in advance.

[0088] 8. The filtered speed data is used to calculate the rate of change of speed, and then input into a preset speed change trend analysis model for trend analysis. Through continuous optimization of this model, using support vector machine algorithms and big data analysis technology, the accuracy of overspeed condition judgment is improved, and the adaptability and robustness of the model are enhanced. The system can more accurately judge the mechanical overspeed condition, avoid the risks caused by inaccurate judgment, and ensure the safe and stable operation of the unit.

[0089] 9. When the mechanical overspeed condition is determined, the optimal switching time and force of the hydraulic switching valve can be calculated based on the rotational inertia of the hydro-generator unit. This can effectively avoid the water hammer effect, ensure rapid mechanical deceleration, and prevent damage to the unit caused by hydraulic system shock due to hydraulic switching valve switching, thereby extending the service life of the unit and hydraulic system.

[0090] 10. During the switching process of the hydraulic switching valve, the pressure change of the hydraulic system is monitored in real time, and a correlation model between the pressure fluctuation and the switching parameter of the hydraulic switching valve is established through a machine learning algorithm. The switching parameter of the hydraulic switching valve can be adjusted according to the pressure fluctuation, and the optimization is iterated until the pressure fluctuation is controlled within a preset range. The closed-loop adjustment mechanism ensures the stability of the hydraulic system and improves the adaptability and reliability of the system under various working conditions.

[0091] 11. The fault self-diagnosis module stores the fault information to the system log and uploads it to the remote monitoring center, analyzes the fault by using natural language processing technology and knowledge graph technology, automatically generates a maintenance work order and distributes it to the relevant maintenance personnel, which facilitates subsequent fault analysis and maintenance work. At the same time, the maintenance results are fed back to update the fault solution knowledge base, which helps to continuously optimize the fault diagnosis and solution process, and improve the maintainability and maintenance efficiency of the entire system. BRIEF DESCRIPTION OF DRAWINGS

[0092] The application will be further described below in conjunction with the drawings and examples:

[0093] Fig. 1 The system structure diagram of the application is shown in the figure;

[0094] Fig. 2 The mechanical overspeed monitoring host structure diagram of the application is shown in the figure;

[0095] Fig. 3 The system operation flowchart of the application is shown in the figure. DETAILED DESCRIPTION

[0096] The preferred scheme is shown in the figure Figs. 1 to 3 A mechanical overspeed monitoring system of a hydroelectric generating set is composed of a fastening ring 1, a main shaft 2, an overspeed pendulum 3, a counterweight 4, a hydraulic switching valve 5, a high-speed camera 6, a support 7, a vibration and swing collection system 8, a rotating speed collection system 9, and a mechanical overspeed monitoring host 10.

[0097] The fastening ring 1 is connected with the main shaft 2 through clamping, and sufficient pre-tightening force is applied to ensure firmness.

[0098] The overspeed pendulum 3 and the counterweight 4 are symmetrically installed on the fastening ring 1, which ensures dynamic balance during rotation and prevents eccentric force.

[0099] The hydraulic switching valve 5 is installed opposite to the overspeed pendulum 3, with a gap of about 5mm, which ensures that the hydraulic switching valve 5 can realize oil path switching during mechanical overspeed.

[0100] The high-speed camera 6 is fixed by the support 7, so that the high-speed camera 6 can include the hydraulic switching valve 5 and the overspeed pendulum 3 in the camera range.

[0101] The swing collection system 8 and the rotation speed collection system 9 are connected with the mechanical overspeed monitoring host 10 through data transmission lines, to ensure that relevant data signals are transmitted.

[0102] The first signal receiving unit 101 of the mechanical overspeed monitoring host 10 is inserted into the lower left side of the mounting rack 105, the second signal receiving unit 102 is inserted into the lower middle side of the mounting rack 105, the power module 103 is inserted into the lower right side of the mounting rack 105, and the display screen 104 is embedded into the top of the mounting rack 105. After the mechanical overspeed monitoring host 10 is powered on, the built-in software can analyze the device motion speed critical value, mechanical hysteresis, and unit swing stability during the mechanical overspeed process.

[0103] A method for operating a mechanical overspeed monitoring system of a hydroelectric generating unit includes the following steps:

[0104] S1, a high-precision rotation speed sensor is used to collect the rotation speed signal of the hydroelectric generating unit in real time, and a filtering algorithm is used to eliminate the influence of electromagnetic interference and mechanical vibration on the signal, to obtain stable rotation speed data.

[0105] A high-precision rotation speed sensor is used to collect the rotation speed signal of the hydroelectric generating unit in real time, to obtain original rotation speed data. A wavelet transform algorithm is used to denoise the collected original rotation speed data, to eliminate the influence of electromagnetic interference on the rotation speed signal. According to the mechanical structure characteristics of the hydroelectric generating unit, a mechanical vibration mathematical model is established, and an adaptive filtering algorithm is used to filter the rotation speed signal, to eliminate the influence of mechanical vibration on the rotation speed signal. The denoised and filtered rotation speed signal is subjected to feature extraction, to obtain time domain and frequency domain characteristic parameters of the rotation speed signal. The extracted rotation speed signal characteristic parameters are input into a support vector machine model, and the model is trained and optimized, to obtain stable and reliable rotation speed estimation values. The estimated rotation speed values are compared with preset rotation speed threshold values, and if the rotation speed exceeds the normal range, it is determined that the hydroelectric generating unit is abnormal, and an alarm signal is triggered. According to the rotation speed estimation values and the alarm signal, an operation state report of the hydroelectric generating unit is generated, to provide data support for equipment maintenance and fault diagnosis.

[0106] Specifically, a high-precision speed sensor with an accuracy of 1% is used to collect the speed signal of the hydro-generator in real time, with 1000 samples per second, to obtain the original speed data. For the collected original speed data, a wavelet transform algorithm db4 wavelet basis function is used to decompose the signal for 5 layers, and then a threshold method is used to denoise the high-frequency components, effectively eliminating the influence of electromagnetic interference on the speed signal. According to the mechanical structure characteristics of the hydro-generator, a mechanical vibration mathematical model considering factors such as bearings, rotors and excitation windings is established, and the speed signal is filtered by a least squares adaptive filtering algorithm to filter out the influence of mechanical vibration on the speed signal. The denoised and filtered speed signal is feature extracted, and the frequency domain feature parameters of the speed signal are obtained by a fast Fourier transform algorithm, including the fundamental frequency, harmonic components, etc., and the time domain feature parameters such as mean and variance are also extracted. The 10 key speed signal feature parameters extracted are input into the support vector machine model, a radial basis kernel function is used, and the model is trained and optimized through cross-validation and grid search, and finally a stable and reliable speed estimation value is obtained, with an estimation accuracy of 2%. The estimated speed value is compared with the preset threshold of rated speed ± 1%, if the speed exceeds the normal range, it is judged that the hydro-generator has an abnormality, and an audible and visual alarm signal is triggered. According to the speed estimation value and the alarm signal, a hydro-generator operation status report containing speed trend chart, alarm log and other contents is automatically generated, providing reliable data support for equipment maintenance and fault diagnosis.

[0107] S2, input the filtered speed data into a preset speed change trend analysis model, if the speed change rate exceeds the preset threshold, it is judged as a mechanical overspeed state.

[0108] Obtain real-time speed data of mechanical equipment, filter the speed data to remove noise interference, and obtain filtered speed data. According to the filtered speed data, calculate the speed change rate to obtain the speed change rate data within a certain time range. Input the speed change rate data into a preset speed change trend analysis model, and analyze the trend of the speed change rate data through the model to determine whether the speed change rate exceeds the preset threshold. If the speed change rate exceeds the preset threshold, it is judged that the mechanical equipment is in an overspeed state, and an overspeed alarm is triggered to notify relevant personnel for processing. If the speed change rate does not exceed the preset threshold, it is judged that the mechanical equipment is in a normal working state, and the real-time speed data is continuously obtained for the next round of speed change trend analysis. According to the historical speed change rate data and the overspeed state judgment result, the speed change trend analysis model is trained and optimized using a support vector machine algorithm to improve the accuracy of the model in judging the overspeed state. Through big data analysis technology, the association rules between the speed change rate and other operating parameters of the mechanical equipment are mined, and the speed change trend analysis model is optimized in combination with the association rules to improve the adaptability and robustness of the model.

[0109] Specifically, after the real-time rotating speed data of the mechanical equipment is collected by the sensor, the rotating speed data is filtered by using the Kalman filtering algorithm, the filtering parameters Q=001 and R=1 are set, and the smoothed rotating speed data is obtained after filtering. According to the filtered rotating speed data, the differential method is used to calculate the rotating speed change rate every 1 second, and the rotating speed change rate data sequence in a period of time is obtained. The rotating speed change rate data sequence is input into the preset LSTM neural network model, and the model can predict the trend of the current rotating speed change rate by training the historical rotating speed change rate data. By comparing the predicted rotating speed change rate with the preset threshold 5, it is judged whether the mechanical equipment is in an overspeed state. If the predicted rotating speed change rate exceeds 5, an overspeed alarm is triggered, and relevant personnel are notified for processing through the ways of short message and email; if the predicted rotating speed change rate does not exceed 5, it is judged that the mechanical equipment is in a normal working state, and the next round of rotating speed change trend analysis is continued. At the same time, by using the historical rotating speed change rate data and the overspeed state judgment result, the support vector machine algorithm is used to optimize the parameters of the LSTM model, the optimal model parameter combination is found by using the grid search method, and the accuracy of the model in judging the overspeed state is improved. In addition, by using the Apriori association rule mining algorithm, the minimum support degree of 05 and the minimum confidence of 8 are used to mine the association rules between the rotating speed change rate and the vibration frequency, temperature and other operating parameters of the mechanical equipment, and it is found that when the rotating speed change rate exceeds 5, the vibration frequency will also exceed 100 Hz, and the temperature will exceed 80 degrees Celsius. According to the mined association rules, the input features of the LSTM model are optimized, and the vibration frequency and temperature are also used as the input of the model, so as to improve the adaptability and robustness of the model. After optimization, the LSTM model can more accurately judge the overspeed state of the mechanical equipment, and effectively avoid the damage and safety accidents caused by overspeed.

[0110] S3, when judging that the mechanical equipment is in an overspeed state, triggering a hydraulic switching valve control module, calculating the switching time and force of the hydraulic switching valve according to the moment of inertia of the hydroelectric generator set, and avoiding water hammer effect.

[0111] The rotational speed of the hydroelectric generator set is monitored in real time by a rotational speed sensor, and the collected rotational speed data is transmitted to a mechanical overspeed monitoring host. After receiving the rotational speed data, the mechanical overspeed monitoring host compares it with a preset mechanical overspeed threshold to determine whether the current is in a mechanical overspeed state. If the result is a mechanical overspeed state, the mechanical overspeed monitoring host immediately sends a trigger signal to a hydraulic switching valve control module. After receiving the trigger signal, the hydraulic switching valve control module obtains the rotational inertia parameter of the hydroelectric generator set from the equipment parameter database. According to the obtained rotational inertia parameter, an optimal switching time and switching force of the hydraulic switching valve are calculated by using an optimization algorithm to achieve rapid and effective mechanical deceleration while avoiding water hammer effect. The calculated switching time and force parameters of the hydraulic switching valve are transmitted to a hydraulic switching valve execution unit to control the hydraulic switching valve to switch according to the optimal scheme. After the hydraulic switching valve is switched, the rotational speed change of the hydroelectric generator set is continuously monitored until the rotational speed returns to the normal range, ensuring that the mechanical overspeed problem is effectively solved.

[0112] Specifically, by installing a high-precision rotational speed sensor on the hydroelectric generator set, rotational speed data is collected in real time, and the sampling frequency of the sensor is set to 1000 Hz, which can meet the real-time monitoring requirements. The rotational speed data is transmitted to the mechanical overspeed monitoring host through industrial Ethernet, and the transmission delay is less than 10 ms. The mechanical overspeed monitoring host sets the mechanical overspeed threshold to 1200 rpm, and uses a real-time comparison algorithm to judge the rotational speed data. When the rotational speed data of three consecutive sampling periods exceeds the threshold, it is determined that the mechanical overspeed is in a state. The mechanical overspeed monitoring host completes the judgment within 2 ms and sends a trigger signal to the hydraulic switching valve control module through a high-speed optical fiber network. The hydraulic switching valve control module uses an equipment parameter management system based on the Internet of Things to query the rotational inertia parameter of the hydroelectric generator set from the database through a unique device ID, and the query time is less than 5 ms. According to the rotational inertia parameter, a fuzzy PID control algorithm is used to calculate the optimal switching time and switching force of the hydraulic switching valve within 10 ms. The switching time is controlled within 100 ms, and the switching force is controlled above 80% of the rated value. The calculation result is transmitted to the hydraulic switching valve execution unit through a high-speed optical fiber network, and the transmission delay is less than 2 ms. The hydraulic switching valve execution unit uses a high-precision servo control system to complete the switching action of the hydraulic switching valve within 5 ms according to the received switching parameters. After switching is completed, the mechanical overspeed monitoring host continuously monitors the rotational speed change at a frequency of 1000 Hz, and uses a Kalman filter algorithm to filter the rotational speed data and predict the rotational speed recovery trend in real time. When the rotational speed data of 10 consecutive sampling periods falls below 1000 rpm, it is determined that the mechanical overspeed problem has been effectively solved, and the entire process is completed within 1 s.

[0113] S4, the hydraulic switching valve control module outputs a control signal to the hydraulic actuator, quickly switches the state of the hydraulic switching valve, and realizes emergency shutdown.

[0114] The hydraulic switching valve control module receives an emergency shutdown instruction, determines whether to perform an emergency shutdown operation, and if so, proceeds to the next step; otherwise, it continues to monitor the emergency shutdown instruction. The hydraulic switching valve control module determines the hydraulic switching valve state that needs to be switched according to the preset hydraulic switching valve switching control strategy, and generates the corresponding hydraulic switching valve control signal. The hydraulic switching valve control module outputs the generated hydraulic switching valve control signal to the hydraulic actuator, and drives the hydraulic actuator to quickly switch the state of the hydraulic switching valve through the control signal. After receiving the control signal output by the hydraulic switching valve control module, the hydraulic actuator quickly switches the state of the hydraulic switching valve according to the control signal, realizing the quick switching of the hydraulic switching valve. The hydraulic switching valve control module monitors the state feedback signal of the hydraulic actuator in real time, determines whether the hydraulic switching valve has been switched to the target state, and if so, proceeds to the next step; otherwise, it continues to output the control signal to drive the hydraulic actuator to switch the state of the hydraulic switching valve. After determining that the hydraulic switching valve has been switched to the target state according to the state feedback signal of the hydraulic actuator, the hydraulic switching valve control module stops outputting the control signal and feeds back the switching result of the hydraulic switching valve state to the upper control system. After receiving the switching result of the hydraulic switching valve state fed back by the hydraulic switching valve control module, the upper control system determines whether the emergency shutdown operation has been completed, and if so, ends the emergency shutdown process; otherwise, it continues to perform other emergency shutdown operations.

[0115] Specifically, after receiving the emergency stop command, the hydraulic switching valve control module determines whether the emergency stop operation needs to be performed by a preset threshold. If the priority of the emergency stop command is higher than the threshold, the control module determines the state of the hydraulic switching valve that needs to be switched according to the preset hydraulic switching valve switching control strategy, determines the state of the hydraulic switching valve that needs to be switched by a fuzzy control algorithm, and generates a 4-20 mA standard current signal output to the hydraulic actuator. After receiving the control signal, the hydraulic actuator quickly adjusts the opening of the hydraulic switching valve by a PID control algorithm to realize the quick switching of the hydraulic switching valve within 100 ms. At the same time, the hydraulic switching valve control module monitors the position feedback signal of the hydraulic actuator in real time by high-speed AD sampling, filters the feedback signal by a Kalman filtering algorithm, and determines whether the hydraulic switching valve has been switched to the target position. If the hydraulic switching valve has not been switched to the target position within 200 ms, the control module continues to output the control signal and adjusts the control parameters in real time by an adaptive control algorithm to ensure that the hydraulic switching valve is quickly and smoothly switched to the target state. When the hydraulic switching valve is switched to the target state, the control module stops outputting the control signal and feeds back the switching result of the hydraulic switching valve state to the upper control system through the industrial Ethernet. The upper control system determines whether the emergency stop operation is completed by a decision tree algorithm according to the feedback result. If not, it continues to perform other emergency stop operations until the entire system is safely shut down.

[0116] S5, during the switching process of the hydraulic switching valve, the pressure change of the hydraulic system is monitored in real time, and if the pressure fluctuation exceeds the preset range, the switching parameters of the hydraulic switching valve are adjusted to ensure the stability of the system.

[0117] Real-time pressure data during the switching process of the hydraulic switching valve is obtained, the pressure data is compared with the preset range, and it is determined whether the pressure fluctuation exceeds the preset range. If the pressure fluctuation exceeds the preset range, the adjustment range of the switching parameters of the hydraulic switching valve is determined according to the degree of the pressure fluctuation exceeding the range. An association model between the pressure fluctuation and the switching parameters of the hydraulic switching valve is established by a machine learning algorithm, and the optimal switching parameters of the hydraulic switching valve are predicted according to the pressure fluctuation. The adjusted switching parameters of the hydraulic switching valve are applied to the switching process of the hydraulic switching valve, and the pressure change is continuously monitored to determine whether the adjusted parameters are effective. If the adjusted parameters fail to effectively reduce the pressure fluctuation, the machine learning model is further optimized, and the optimal switching parameters of the hydraulic switching valve are re-predicted. The switching parameters of the hydraulic switching valve are continuously iteratively optimized until the pressure fluctuation is controlled within the preset range, ensuring the stable operation of the hydraulic system. The optimized switching parameters of the hydraulic switching valve are saved as preset parameters, which are directly applied in the subsequent switching process of the hydraulic switching valve, improving the response speed and stability of the system.

[0118] Specifically, during the switching process of the hydraulic switching valve, pressure data is collected in real time by a pressure sensor with a sampling frequency of 1000 Hz, and the collected pressure data is compared with a preset pressure range, which is 5-0 MPa. When it is detected that the pressure fluctuation exceeds the preset range, the adjustment range of the switching parameter of the hydraulic switching valve is determined according to the degree of exceeding the range by using a fuzzy control algorithm. The greater the pressure fluctuation, the greater the adjustment range, and the adjustment range is 1-5 MPa. At the same time, a correlation model between pressure fluctuation and the switching parameter of the hydraulic switching valve is established by using historical pressure fluctuation data and the corresponding switching parameter of the hydraulic switching valve through a support vector machine algorithm, and the optimal switching parameter of the hydraulic switching valve is predicted according to the current pressure fluctuation. The optimized switching parameter of the hydraulic switching valve is applied to the switching process of the hydraulic switching valve, and the pressure change is continuously monitored at a period of 10 ms, and the pressure fluctuation is analyzed by statistical method to determine whether the adjusted parameter is effective. If the adjusted parameter fails to control the pressure fluctuation within the preset range, the machine learning model is optimized by using an incremental learning algorithm, and a genetic algorithm is introduced for parameter optimization to re-predict the optimal switching parameter of the hydraulic switching valve. Through continuous iterative optimization, the pressure fluctuation is finally controlled within the preset range of 5-0 MPa, ensuring the stable operation of the hydraulic system. The optimized switching parameter of the hydraulic switching valve is saved as a preset parameter, which can be directly called in the subsequent switching process of the hydraulic switching valve, and the system response time can be shortened to within 50 ms, significantly improving the real-time performance and stability of the system.

[0119] S6, the system is built-in fault self-diagnosis module, regularly detect the working state of the speed sensor, hydraulic switching valve control module and signal processing module, if detected fault, trigger fault lock function, prevent false alarm.

[0120] According to a preset time interval, a built-in fault self-diagnosis module of the system is triggered regularly to detect the state of the rotating speed sensor, the hydraulic switching valve control module and the signal processing module. When detecting the state, the fault self-diagnosis module acquires real-time rotating speed data of the rotating speed sensor and compares the data with a preset normal rotating speed range to determine whether the rotating speed sensor has a fault. Meanwhile, the fault self-diagnosis module analyzes feedback signals of the hydraulic switching valve control module to determine whether the opening and closing state of the hydraulic switching valve is consistent with the control command and determine whether the hydraulic switching valve control module has a fault. The fault self-diagnosis module also monitors input and output data of the signal processing module in real time, analyzes the continuity and integrity of the data, and determines whether the signal processing module works normally. If the fault self-diagnosis module detects that any module has a fault, the fault self-locking function is triggered immediately to stop the work of the corresponding module by cutting off the power supply or blocking the signal transmission. At the same time of triggering the fault self-locking function, the fault self-diagnosis module generates a fault report, records the time, position and fault type and the like of the fault occurrence, and sends the report to the system administrator for subsequent maintenance. The fault self-locking function continues to take effect until the system administrator confirms that the fault has been eliminated and manually releases the locking state, and the system can resume normal work, thereby effectively preventing false alarm caused by the continued operation of the faulty module.

[0121] Specifically, the built-in fault self-diagnosis module of the system is automatically triggered every 30 minutes to comprehensively detect the state of the rotating speed sensor, the hydraulic switching valve control module and the signal processing module. The fault self-diagnosis module acquires the rotating speed data of the rotating speed sensor in real time through the RS-485 bus and compares the data with the preset normal rotating speed range of 800-1200 rpm. If the actual rotating speed exceeds the range, it is determined that the rotating speed sensor has a fault. Meanwhile, the fault self-diagnosis module analyzes the feedback signals of the hydraulic switching valve control module based on the feedback signal analysis method based on the PID control algorithm, calculates the deviation value of the control command and the actual execution state, and if the deviation value is greater than 5%, it is determined that the hydraulic switching valve control module has a fault. In addition, the fault self-diagnosis module also monitors the input and output data of the signal processing module in real time by using the CRC check algorithm, analyzes the continuity and integrity of the data frames, and if it is found that there are check errors in three consecutive data frames, it is determined that the signal processing module works abnormally. Once the fault self-diagnosis module detects that any module has a fault, the fault self-locking function is triggered immediately to cut off the power supply of the fault module through the relay, and a fault report containing the fault time, position and type and the like is sent to the system administrator. Before the system administrator manually releases the locking state, the fault module will be in a shutdown state, thereby avoiding the false alarm caused by the continued operation of the fault module.

[0122] S7, the fault self-diagnosis module stores the fault information to the system log, and uploads to the remote monitoring center through the communication module, facilitating subsequent analysis and maintenance.

[0123] According to the fault information obtained by the fault self-diagnosis module, the fault information is written into the system log to obtain a system log file containing the fault information. A communication connection is established with the remote monitoring center through the communication module, and if the communication connection is successfully established, the system log file is uploaded to the remote monitoring center. After receiving the uploaded system log file in the remote monitoring center, the system log file is analyzed using natural language processing technology, and the key fault information is extracted. According to the extracted fault information, semantic analysis is performed through knowledge graph technology to determine the fault type and fault cause, and obtain the fault diagnosis result. The fault diagnosis result is matched with the pre-established fault solution knowledge base to obtain the corresponding fault solution. According to the obtained fault solution, a fault maintenance work order is automatically generated, and the maintenance work order is assigned to the relevant maintenance personnel, notifying the maintenance personnel to perform fault maintenance. After completing the fault maintenance, the maintenance personnel feed back the maintenance result to the remote monitoring center, and the remote monitoring center updates the fault solution knowledge base to optimize the fault diagnosis and solution process.

[0124] Specifically, the fault self-diagnosis module obtains relevant information when the system fails through real-time monitoring of the system's operating state, such as the fault occurrence time of May 10, 2023, 14:30:25, the fault code of ERR_0x001, and the fault description of "system cannot start normally". These fault information is written into the system log file system_log_2023051txt. The communication module uses TCP / IP protocol to establish communication connection with remote monitoring center, and through three times of handshaking mechanism to ensure the reliability of connection. After the connection is established successfully, the system log file is uploaded to the remote monitoring center through FTP protocol. The remote monitoring center uses a natural language processing model based on deep learning to analyze the received system log file, and extracts the fault occurrence time, fault code, fault description and other key information through part-of-speech tagging, named entity recognition and other technologies. Then, by using the pre-constructed fault knowledge graph and through semantic analysis technology, it is judged that the fault type is "system startup failure" and the fault reason is "configuration file damage". According to the fault type and reason, the corresponding solution in the fault solution knowledge base is searched as "reconfigure system parameters and repair damaged configuration file". The system automatically generates a fault maintenance work order M_20230510001 and assigns it to the system administrator Zhang San, and notifies Zhang San to maintain the fault through SMS and email. Zhang San reconfigures and repairs the system according to the maintenance work order, and after completing the maintenance, he feeds back the maintenance result "system has returned to normal operation" to the remote monitoring center. According to the feedback of the maintenance result, the remote monitoring center optimizes the fault solution in the knowledge base, improves the effectiveness of the solution from 85% to 95%, and fine-tunes the fault diagnosis model, improving the accuracy of fault diagnosis. Through this series of intelligent fault diagnosis and maintenance process, the reliability and maintenance efficiency of the system are greatly improved, and the demand for manual intervention is reduced.

[0125] S8、System adopts modular design, each functional module has waterproof, dustproof, anticorrosion ability, adapts to wide range of temperature change, supports plug and play function, is convenient for installation and maintenance.

[0126] According to the modular design of the system, the waterproof, dustproof and corrosion-resistant attribute parameters of each functional module are input into the module attribute database. By querying the module attribute database, the waterproof, dustproof and corrosion-resistant level parameters of each functional module are obtained. According to the obtained waterproof, dustproof and corrosion-resistant level parameters, the fuzzy comprehensive evaluation method is adopted to calculate the comprehensive protection level of each functional module. The comprehensive protection level of each functional module is compared with the preset protection level threshold value, and if the comprehensive protection level is greater than or equal to the threshold value, it is judged that the module meets the system protection requirements. According to the temperature adaptation attribute, the working temperature range parameters of each functional module are obtained and input into the module temperature adaptation database. By querying the module temperature adaptation database, the maximum working temperature and the minimum working temperature of all modules of the system are obtained. According to the obtained maximum working temperature and minimum working temperature, the working temperature range of the system as a whole is determined, and whether the temperature range meets the system temperature adaptation requirements is judged.

[0127] Specifically, according to the modular design of the system, the waterproof, dustproof and corrosion-resistant attribute parameters of each functional module are input into the module attribute database. For example, the waterproof level of module A is IPX7, the dustproof level is IP6X, and the corrosion-resistant level is H; the waterproof level of module B is IPX5, the dustproof level is IP5X, and the corrosion-resistant level is M. By querying the module attribute database through SQL statements, the waterproof, dustproof and corrosion-resistant level parameters of each functional module are obtained. According to the obtained waterproof, dustproof and corrosion-resistant level parameters, the fuzzy comprehensive evaluation method is adopted, the weight vector A=(3, 3, 4) is set, and the comprehensive protection level of each functional module is calculated by weighted average, such as the comprehensive protection level of module A is 85, and the comprehensive protection level of module B is 76. The comprehensive protection level of each functional module is compared with the preset protection level threshold value 8, module A meets the system protection requirements, and module B does not meet the requirements. According to the temperature adaptation attribute, the working temperature range parameters of each functional module are obtained and input into the module temperature adaptation database, such as the working temperature range of module A is -20℃~60℃, and the working temperature range of module B is -10℃~50℃. By querying the module temperature adaptation database, the maximum working temperature of all modules of the system is obtained, which is 60℃, and the minimum working temperature is -20℃. According to the obtained maximum working temperature and minimum working temperature, the working temperature range of the system as a whole is determined to be -20℃~60℃, and whether the temperature range meets the system temperature adaptation requirements is judged, i.e. whether it is within the preset range of -30℃~70℃, which meets the requirements in this example. In summary, through the modular design of the system, the protection and temperature adaptation attribute parameters of each module are parameterized, which can quantitatively evaluate the comprehensive protection level and the adaptive temperature range of the system, providing quantitative basis for system design.

[0128] The above embodiments are only preferred technical solutions of the present application, and should not be regarded as a limitation of the present application. The protection scope of the present application should be the technical solutions recited in the claims, including equivalent replacement solutions of the technical features recited in the claims. That is, equivalent replacement improvements within this scope are also within the protection scope of the present application.

Claims

1. A method of operating a mechanical overspeed monitoring system for a hydroelectric generating unit, the method comprising: The method comprises the following steps: ​ Real-time collection of the rotating speed signal of the hydroelectric generator set by using a high-precision rotating speed sensor, elimination of the influence of electromagnetic interference and mechanical vibration on the signal by using a filtering algorithm, and obtaining of stable rotating speed data; Input of the filtered rotating speed data into a preset rotating speed change trend analysis model, and judgment of a mechanical overspeed state if the rotating speed change rate exceeds a preset threshold value; Triggering of a hydraulic switching valve control module when the mechanical overspeed state is judged, calculation of the switching time and strength of the hydraulic switching valve according to the moment of inertia of the hydroelectric generator set, and avoidance of the water hammer effect; Output of a control signal from the hydraulic switching valve control module to a hydraulic actuator, rapid switching of the state of the hydraulic switching valve, and realization of emergency shutdown; Real-time monitoring of the pressure change of the hydraulic system during the switching process of the hydraulic switching valve, adjustment of the switching parameters of the hydraulic switching valve to ensure the stability of the system if the pressure fluctuation exceeds a preset range; Built-in fault self-diagnosis module of the system, periodic detection of the working state of the rotating speed sensor, the hydraulic switching valve control module and the signal processing module, triggering of a fault locking function to prevent false alarms if a fault is detected; Storage of fault information in the system log by the fault self-diagnosis module, and uploading of the fault information to a remote monitoring center through a communication module for subsequent analysis and maintenance.

2. The method of claim 1, wherein The method comprises the following steps: Real-time collection of the rotating speed signal of the hydroelectric generator set by using a high-precision rotating speed sensor, elimination of the influence of electromagnetic interference and mechanical vibration on the signal by using a filtering algorithm, and obtaining of stable rotating speed data, including: Real-time collection of the rotating speed signal of the hydroelectric generator set by using a high-precision rotating speed sensor, and acquisition of original rotating speed data; Denoising processing of the collected original rotating speed data by using a wavelet transform algorithm, and elimination of the influence of electromagnetic interference on the rotating speed signal; Establishment of a mechanical vibration mathematical model according to the mechanical structure characteristics of the hydroelectric generator set, filtering of the rotating speed signal by using an adaptive filtering algorithm, and elimination of the influence of mechanical vibration on the rotating speed signal; Feature extraction of the denoised and filtered rotating speed signal, and acquisition of time domain and frequency domain feature parameters of the rotating speed signal; Input of the extracted rotating speed signal feature parameters into a support vector machine model, model training and optimization, and obtaining of stable and reliable rotating speed estimation values; Comparison of the estimated rotating speed values with preset rotating speed threshold values, judgment of the existence of an abnormality of the hydroelectric generator set if the rotating speed exceeds the normal range, and triggering of an alarm signal; Generation of an operation state report of the hydroelectric generator set according to the rotating speed estimation values and the alarm signal, and provision of data support for equipment maintenance and fault diagnosis.

3. The method of claim 1, wherein the method further comprises: The method comprises the following steps: Input of the filtered rotating speed data into a preset rotating speed change trend analysis model, and judgment of a mechanical overspeed state if the rotating speed change rate exceeds a preset threshold value, including: Acquisition of real-time rotating speed data of a mechanical device, filtering processing of the rotating speed data to remove noise interference in the rotating speed data, and obtaining of filtered rotating speed data; Calculation of the rotating speed change rate according to the filtered rotating speed data, and obtaining of rotating speed change rate data within a certain time range; Input of the rotating speed change rate data into a preset rotating speed change trend analysis model, trend analysis of the rotating speed change rate data by using the model, and judgment of whether the rotating speed change rate exceeds a preset threshold value; If the rotational speed change rate exceeds the preset threshold, it is judged that the mechanical equipment is in an overspeed state, an overspeed alarm is triggered, and relevant personnel are notified to handle it; If the rotational speed change rate does not exceed the preset threshold, it is judged that the mechanical equipment is in a normal working state, real-time rotational speed data is continuously acquired, and the next round of rotational speed change trend analysis is performed; According to the historical rotational speed change rate data and the overspeed state judgment result, the support vector machine algorithm is used to train and optimize the rotational speed change trend analysis model, so as to improve the accuracy of the model in judging the overspeed state; Through big data analysis technology, the association rules between the rotational speed change rate and other operating parameters of the mechanical equipment are mined, and the rotational speed change trend analysis model is optimized combined with the association rules, so as to improve the adaptability and robustness of the model.

4. The method of claim 1, wherein The steps include: When it is judged that the mechanical equipment is in an overspeed state, a hydraulic switching valve control module is triggered, the switching time and force of the hydraulic switching valve are calculated according to the moment of inertia of the hydroelectric generator set, and the water hammer effect is avoided, including: The rotational speed of the hydroelectric generator set is monitored in real time through a rotational speed sensor, and the collected rotational speed data is transmitted to a mechanical overspeed monitoring host; After receiving the rotational speed data, the mechanical overspeed monitoring host compares it with a preset mechanical overspeed threshold to judge whether the current is in a mechanical overspeed state; If the judgment result is a mechanical overspeed state, the mechanical overspeed monitoring host immediately sends a trigger signal to the hydraulic switching valve control module; After receiving the trigger signal, the hydraulic switching valve control module obtains the moment of inertia parameter of the hydroelectric generator set from a device parameter database; According to the obtained moment of inertia parameter, an optimal switching time and switching force of the hydraulic switching valve are calculated by using an optimization algorithm, so as to realize fast and effective mechanical deceleration while avoiding the water hammer effect; The calculated hydraulic switching valve switching time and force parameters are transmitted to a hydraulic switching valve execution unit to control the hydraulic switching valve to switch according to the optimal scheme; After the hydraulic switching valve is switched, the rotational speed change of the hydroelectric generator set is continuously monitored until the rotational speed returns to the normal range, so as to ensure that the mechanical overspeed problem is effectively solved.

5. The method of claim 1, wherein the method further comprises: determining if the speed of the generator exceeds the predetermined speed threshold; and if the speed of the generator exceeds the predetermined speed threshold, generating an alarm signal. The steps include: the hydraulic switching valve control module outputs a control signal to the hydraulic actuator to quickly switch the state of the hydraulic switching valve and realize emergency shutdown, including: The hydraulic switching valve control module receives an emergency shutdown instruction and judges whether the emergency shutdown operation needs to be performed. If the emergency shutdown operation needs to be performed, the next step is entered; Otherwise, the emergency shutdown instruction is continuously monitored; The hydraulic switching valve control module determines the state of the hydraulic switching valve that needs to be switched according to a preset hydraulic switching valve switching control strategy, and generates a corresponding hydraulic switching valve control signal; The hydraulic switching valve control module outputs the generated hydraulic switching valve control signal to the hydraulic actuator, and drives the hydraulic actuator to quickly switch the state of the hydraulic switching valve through the control signal; After receiving the control signal output by the hydraulic switching valve control module, the hydraulic actuator quickly switches the state of the hydraulic switching valve according to the control signal to realize the quick opening and closing of the hydraulic switching valve; The hydraulic switching valve control module monitors the state feedback signal of the hydraulic actuator in real time, judges whether the hydraulic switching valve has switched to the target state, if it has switched to the target state, then enter the next step; Otherwise, continue to output the control signal to drive the hydraulic actuator to switch the state of the hydraulic switching valve; After the hydraulic switching valve control module determines that the hydraulic switching valve has switched to the target state according to the state feedback signal of the hydraulic actuator, it stops outputting the control signal and feeds back the switching result of the hydraulic switching valve state to the upper control system; After the upper control system receives the switching result of the hydraulic switching valve state fed back by the hydraulic switching valve control module, it judges whether the emergency stop operation has been completed, if it has been completed, then end the emergency stop process; Otherwise, continue to perform other emergency stop operations.

6. The method of claim 1, wherein The steps include: During the switching process of the hydraulic switching valve, the pressure change of the hydraulic system is monitored in real time, if the pressure fluctuation exceeds the preset range, then adjust the switching parameters of the hydraulic switching valve to ensure system stability, including: Obtain real-time pressure data during the switching process of the hydraulic switching valve, compare the pressure data with the preset range, and judge whether the pressure fluctuation exceeds the preset range; If the pressure fluctuation exceeds the preset range, then determine the adjustment range of the switching parameters of the hydraulic switching valve according to the degree of the pressure fluctuation exceeding the range; Establish a correlation model between pressure fluctuation and switching parameters of the hydraulic switching valve through machine learning algorithm, predict the optimal switching parameters of the hydraulic switching valve according to the pressure fluctuation; Apply the adjusted switching parameters of the hydraulic switching valve to the switching process of the hydraulic switching valve, and continuously monitor the pressure change to determine whether the adjusted parameters are effective; If the adjusted parameters fail to effectively reduce the pressure fluctuation, then further optimize the machine learning model and re-predict the optimal switching parameters of the hydraulic switching valve; Continuously optimize the switching parameters of the hydraulic switching valve until the pressure fluctuation is controlled within the preset range to ensure stable operation of the hydraulic system; Save the optimized switching parameters of the hydraulic switching valve as preset parameters, which can be directly applied in subsequent switching process of the hydraulic switching valve, improving the system response speed and stability.

7. The method of claim 1, wherein the method further comprises: determining a mechanical overspeed condition of the hydroelectric generator set based on the comparison of the measured speed to the predetermined speed threshold. The steps include: the system is built-in with a fault self-diagnosis module, which periodically detects the working state of the speed sensor, the hydraulic switching valve control module and the signal processing module, if a fault is detected, then trigger the fault lock function to prevent false alarms, including: ​ Periodically trigger the system's built-in fault self-diagnosis module according to the preset time interval to detect the state of the speed sensor, the hydraulic switching valve control module and the signal processing module; When detecting the state, the fault self-diagnosis module obtains the real-time speed data of the speed sensor and compares it with the preset normal speed range to determine whether the speed sensor has a fault; At the same time, the fault self-diagnosis module analyzes the feedback signal of the hydraulic switching valve control module to determine whether the opening and closing state of the hydraulic switching valve is consistent with the control command to determine whether the hydraulic switching valve control module has a fault; The fault self-diagnosis module also monitors the input and output data of the signal processing module in real time, analyzes the continuity and integrity of the data to determine whether the signal processing module is working normally; If the fault self-diagnosis module detects any fault in any module, the fault lockout function is triggered immediately to stop the work of the corresponding module by cutting off the power supply or blocking the signal transmission; At the same time of triggering the fault lockout function, the fault self-diagnosis module generates a fault report, records the time and location of the fault occurrence, the fault type and other information, and sends the report to the system administrator for subsequent maintenance; The fault lockout function continues to take effect until the system administrator confirms that the fault has been eliminated and manually releases the lockout state, and the system can resume normal work, thereby effectively preventing false alarm caused by the continued operation of the faulty module.

8. The method of claim 1, wherein The steps include: The fault self-diagnosis module stores the fault information in the system log and uploads it to the remote monitoring center through the communication module for subsequent analysis and maintenance, including: According to the fault information obtained by the fault self-diagnosis module, the fault information is written into the system log to obtain a system log file containing the fault information; Through the communication module, a communication connection is established with the remote monitoring center, and if the communication connection is successfully established, the system log file is uploaded to the remote monitoring center; After the remote monitoring center receives the uploaded system log file, the system log file is analyzed using natural language processing technology to extract key fault information; According to the extracted fault information, semantic analysis is performed through knowledge graph technology to determine the fault type and fault cause, and a fault diagnosis result is obtained; The fault diagnosis result is matched with the pre-established fault solution knowledge base to obtain the corresponding fault solution; According to the obtained fault solution, a fault maintenance work order is automatically generated, and the maintenance work order is assigned to the relevant maintenance personnel to notify them to perform fault maintenance; After completing the fault maintenance, the maintenance personnel feed back the maintenance result to the remote monitoring center, and the remote monitoring center updates the fault solution knowledge base to optimize the fault diagnosis and solution process.

9. The method of claim 1, wherein the method further comprises: A mechanical overspeed monitoring system for a hydro-generator unit is adopted, which comprises a fastening ring (1), the fastening ring (1) is connected with a main shaft (2), an overspeed pendulum (3) and a counterweight (4) are installed on the fastening ring (1), the overspeed pendulum (3) is oppositely installed with a hydraulic switching valve (5), the state of the overspeed pendulum (3) and the hydraulic switching valve (5) is monitored by a high-speed camera (6); The hydraulic switching valve (5) and the high-speed camera (6) are electrically connected with a mechanical overspeed monitoring host (10), the mechanical overspeed monitoring host (10) is electrically connected with a vibration and swing acquisition system (8) and a rotating speed acquisition system (9).

10. The method of claim 9, wherein the method further comprises: The fastening ring (1) is connected with the main shaft (2) by clamping.

11. The method of claim 9, wherein the method further comprises: The overspeed pendulum (3) and the counterweight (4) are symmetrically installed on the fastening ring (1).

12. The method of claim 9, wherein the method further comprises: The high-speed camera (6) is fixed by a support (7), so that the high-speed camera (6) can take the whole hydraulic switching valve (5) and the overspeed pendulum (3) into the camera range.

13. The method of claim 9, wherein the method further comprises: The vibration and swing acquisition system (8) and the rotating speed acquisition system (9) are connected with the mechanical overspeed monitoring host (10) through data transmission lines to ensure that the relevant data signals are transmitted.

14. The method of claim 9, wherein the method further comprises: The mechanical overspeed monitoring host (10) comprises a first signal receiving unit (101), a second signal receiving unit (102), a power module (103), a display screen (104) and a rack (105), the signal receiving unit (101) is inserted and installed at the left lower part of the mounting rack (105), the second signal receiving unit (102) is inserted and installed at the middle lower part of the mounting rack (105), the power module (103) is inserted and installed at the right lower part of the mounting rack (105), and the display screen (104) is embedded and installed at the top of the rack (105).

Images