Intelligent combined control method applied to drainage system of hydropower station

Abstract

The invention discloses an intelligent combined control method applied to a hydropower station drainage system, and the method comprises the steps: carrying out the calculation of the drainage demands of a hydropower station, and dividing the drainage demands of the hydropower station into low, medium and high drainage demands; according to a fuzzy rule, outputting membership degrees of low, medium and high drainage demands and a comprehensive drainage demand; the start-stop alternating logic of the pump is constructed based on the membership degrees of the low, medium and high drainage requirements; performing fault analysis on the running state of the pump; according to the method, starting, stopping and operation parameters of the three drainage system devices can be automatically adjusted according to the actual operation condition of the hydropower station, so that the purposes that normal operation and safety of a unit are guaranteed, safe maintenance of the unit is facilitated, meanwhile, it is guaranteed that the interior of a plant is not accumulated with water or moist, and the safety and economical efficiency of the hydropower station are improved are achieved.

Description

Technical Field

[0001] The invention belongs to the technical field of drainage of hydropower stations, and in particular relates to an intelligent joint control method applied to a drainage system of a hydropower station. Background Art

[0002] At present, the drainage system of the hydropower station mainly includes the maintenance drainage of the hydropower station unit, the leakage drainage system and the gallery pumping system; the unit maintenance drainage system is to discharge the water in the pressure steel pipe to the downstream when the unit is shut down for maintenance, so as to carry out safe maintenance of the unit; the system is composed of a maintenance valve, a drainage pipe, a drainage gate, etc. The maintenance valve is installed at the upstream end of the pressure steel pipe, the drainage pipe connects the maintenance valve and the drainage gate, and the drainage gate is installed at the downstream end of the pressure steel pipe. By opening and closing the maintenance valve and the drainage gate, the water level in the pressure steel pipe is controlled; the leakage drainage system is to discharge the leakage water of the pressure steel pipe, the volute, the guide vane and other components to the downstream when the unit is running, so as to ensure the normal operation and safety of the unit; the system is composed of a leakage valve, a leakage pipe, a leakage gate, etc. The leakage valve is installed at the leakage port of each component, the leakage pipe connects the leakage valve and the leakage gate, and the leakage gate is installed on the outer wall of the plant. By opening and closing the leakage valve and the leakage gate, the leakage amount of each component is controlled. The corridor drainage system is used to pump out water on the ground and water from equipment leakage in the factory when the unit is running or shut down, so as to keep the factory clean and dry. The system consists of a drainage pump, drainage pipes, drainage gates, etc. The drainage pump is installed in the basement of the factory or below the ground. The drainage pipe connects the drainage pump and the drainage gate. The drainage gate is installed on the outer wall of the factory. The amount of water accumulation in the factory is controlled by opening or closing the drainage pump and the drainage gate.

[0003] At present, the operation logic of the drainage system of the hydropower station is: set the start and stop water levels of each drainage system in advance, each drainage pump works when the water level reaches the start water level, stops when the water level is pumped below the set stop water level, and multiple drainage pumps rotate according to the number of starts and stops. The operation logic of the conventional drainage system is relatively simple. When the water level in the water collection well rises rapidly, it cannot accurately determine the number of drainage pumps to start according to the water level changes; and the operation characteristics of each drainage pump motor are also different. The current drainage pump cannot change the operation logic in real time according to the current state of the motor.

[0004] Therefore, it is necessary to design an intelligent joint control method for the drainage system of a hydropower station to solve the above problems. Summary of the invention

[0005] The technical problem to be solved by the present invention is to provide an intelligent joint control method applied to the drainage system of a hydropower station. The method can automatically adjust the start and stop and operating parameters of the above three drainage system equipment according to the actual operation of the hydropower station, so as to ensure the normal operation and safety of the unit and facilitate safe maintenance of the unit, while ensuring that there is no water accumulation and moisture inside the plant, thereby improving the safety and economy of the hydropower station.

[0006] In order to achieve the above technical effects, the technical solution adopted by the present invention is: An intelligent joint control method applied to a drainage system of a hydropower station comprises the following steps: S1, calculate the drainage demand of the hydropower station, divide the drainage demand of the hydropower station into low, medium and high drainage demand; output the membership degree and comprehensive drainage demand of low, medium and high drainage demand according to the fuzzy rules; S2, constructs the pump start-stop rotation logic based on the membership of low, medium and high drainage demands; S3, perform fault analysis on the operating status of the pump.

[0007] Furthermore, when calculating drainage demand, sensors installed at various drainage points are used to monitor the relevant data on the hydropower station's leakage, maintenance drainage, production water drainage and water accumulation in the plant area in real time; the relevant data include drainage flow data and accumulated water level data, and the data is transmitted to the computer system in the central control room.

[0008] Preferably, in step S1, the specific method for calculating the drainage demand of the hydropower station is: S101, assuming that the water level and flow rate ranges from [0, P] meters and [0, N] cubic meters per second, respectively, and the drainage demand ranges from [0, 1], where 0 indicates no drainage is required and 1 indicates the maximum drainage is required; Set low, medium and high thresholds for water level and flow respectively; divide water level and flow into several fuzzy subsets, including low, medium and high categories; S102, setting a membership function for each fuzzy subset to describe its membership degree: The membership functions of water level include: The membership function of the low water level is A_L(x)=(Px) / P,x∈[0,P]; The membership function of the middle water level is A_M(x)=min(x / (P / 2),(Px) / (P / 2)),x∈[0,P]; The membership function of the high water level is A_H(x)=x / P,x∈[0,P]; The membership functions of traffic include: The membership function of low flow is A_S(y)=(Ny) / N,y∈[0,N]; The membership function of medium flow is A_M(y)=min(y / (N / 2),(Ny) / (N / 2)),y∈[0,N]; The membership function of high flow is A_L(y)=y / N,y∈[0,N]; S103, establishing fuzzy rules for water level, flow and drainage demand based on binary comparison ranking method; S104, obtaining output data through fuzzy reasoning and defuzzification; fuzzy reasoning calculates the membership of the output data according to the membership of the input data and the fuzzy rules; according to the fuzzy rules, the output data belongs to the membership of low, medium and high drainage demands.

[0009] Preferably, in step S104, according to the fuzzy rule, the specific method of outputting the membership of the data to low, medium and high drainage demands is: Low drainage demand membership: B_L=min(A_L(x),A_S(y)); Degree of subordination of drainage demand: B_M=max(min(A_L(x),A_M(y)),min(A_M(x),A_S(y)),min(A_M(x),A_M(y))); High drainage demand membership: B_H=max(min(A_L(x),A_L(y)),min(A_M(x),A_L(y)),min(A_H(x),A_S(y)),min(A_H(x),A_M(y)),min(A_H(x),A_L(y))); The comprehensive drainage demand is calculated by the centroid method: Y=(B_L*0+B_M*0.5+B_H*1) / (B_L+B_M+B_H); In the formula, x represents water level data and y represents flow data.

[0010] Preferably, in step S2, the specific method of constructing the start-stop rotation logic of the pump based on the membership of low, medium and high drainage demands is: S201, collecting historical drainage data of the drainage system and setting the priority of the drainage system: The drainage system includes the maintenance drainage system, the leakage drainage system and the corridor drainage system. The priorities of the drainage systems are set to F1, F2 and F3 according to the average value of the historical drainage volume; the drainage system with the smallest drainage volume is set to the highest priority; S202, setting the maximum drainage threshold of the drainage system with F1 priority to Q1, the maximum drainage threshold of the drainage system with F2 priority to Q2, and the maximum drainage threshold of the drainage system with F3 priority to Q3; S203, setting a drainage allocation strategy based on the comprehensive drainage demand Y: The drainage system with F1 priority is allocated for drainage first, the drainage demand exceeding Q1 is allocated to the drainage system with F2 priority, and the drainage demand exceeding Q2 is allocated to the drainage system with F3 priority; S204, collecting the working range and performance curve data of each pump in the maintenance drainage system, and determining the optimal drainage volume Q' of the pump from the performance curve; S205, numbering each pump, the numbered numbers forming an array Z; randomly sorting the numbered numbers in the array Z by a random algorithm to form a sorted pump sequence; S206, after receiving the allocated drainage demand, the drainage system starts the pumps in sequence according to the pump sequence in S205 to drain the water, and controls the power through the frequency converter of the pump. When the pump reaches the optimal drainage volume Q', it stops increasing the power and starts the next sequenced pump, and so on, until the drainage demand allocated by the drainage system can be matched; S207, when the drainage demand Y exceeds Q3, the controller synchronously increases the power of all the pumps started in the three drainage systems, evenly increases the power of all the pumps in use, so that the total drainage volume matches the comprehensive drainage demand Y; the drainage volume of each pump does not exceed its working range.

[0011] Furthermore, the random algorithm is implemented by using the Knuth shuffle algorithm to randomly arrange multiple numbered numbers; each time the drainage demand is redistributed, the start and stop sequence of the pumps is rearranged through the random algorithm.

[0012] Preferably, the maximum drainage thresholds Q1, Q2 and Q3 do not exceed the sum of the optimal drainage volumes Q' of all pumps in the corresponding drainage system.

[0013] Preferably, performing fault analysis on the operating status of the pump includes: The feedback signals of each drainage system are received through the computer system, including the real-time drainage data and power data of each pump; the fault diagnosis is performed on the collected data as follows: Data preprocessing: Clean, standardize and reduce noise of collected data to improve data quality and availability; Feature extraction: Use machine learning algorithms to extract features from preprocessed data and extract feature vectors that reflect the operating status of the equipment; Fault diagnosis: Use artificial intelligence algorithms such as classification, clustering, or regression to diagnose faults on feature vectors to determine whether the equipment has a fault and the type and extent of the fault.

[0014] Fault alarm: Based on the results of fault diagnosis, an alarm will be issued for the faulty equipment, including information on the fault location, time and cause, and corresponding maintenance suggestions will be provided.

[0015] The beneficial effects of the present invention are as follows: 1. The present invention proposes an intelligent joint control method for the drainage system of a hydropower station. The method can automatically adjust the start and stop and operation parameters of the above three drainage system equipment according to the actual operation of the hydropower station, so as to ensure the normal operation and safety of the unit and facilitate the safe maintenance of the unit, while ensuring that there is no water accumulation and moisture inside the plant, thereby improving the safety and economy of the hydropower station.

[0016] 2. The present invention can reduce the problem of frequent start and stop of drainage pumps and long-term no-load or overload operation that causes easy damage to the equipment. The drainage pump can be put into operation for a long time using this method. Compared with the existing scheme, the trend of water outflow changes can be analyzed in advance, and the drainage pump can be started in advance when the water inflow increases, so as to drain the accumulated water as soon as possible and ensure the safe and stable operation of the hydropower station. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 It is a schematic diagram of the process in the present invention; Figure 2 It is a schematic diagram of the start-stop rotation logic block diagram in the present invention; Figure 3 This is a schematic diagram of fault diagnosis in the present invention; Figure 4 It is a schematic diagram of a specific case of fault diagnosis in the second embodiment of the present invention. DETAILED DESCRIPTION

[0018] Embodiment 1: like Figure 1 As shown, an intelligent joint control method applied to a drainage system of a hydropower station comprises the following steps: S1, calculate the drainage demand of the hydropower station, divide the drainage demand of the hydropower station into low, medium and high drainage demand; output the membership degree and comprehensive drainage demand of low, medium and high drainage demand according to the fuzzy rules; S2, constructs the pump start-stop rotation logic based on the membership of low, medium and high drainage demands; S3, perform fault analysis on the operating status of the pump.

[0019] Furthermore, when calculating drainage demand, sensors installed at various drainage points are used to monitor the relevant data on the hydropower station's leakage, maintenance drainage, production water drainage and water accumulation in the plant area in real time; the relevant data include drainage flow data and accumulated water level data, and the data is transmitted to the computer system in the central control room.

[0020] Preferably, in step S1, the specific method for calculating the drainage demand of the hydropower station is: S101, assuming that the water level and flow rate ranges from [0, P] meters and [0, N] cubic meters per second, respectively, and the drainage demand ranges from [0, 1], where 0 indicates no drainage is required and 1 indicates the maximum drainage is required; Set low, medium and high thresholds for water level and flow respectively; divide water level and flow into several fuzzy subsets, including low, medium and high categories; S102, setting a membership function for each fuzzy subset to describe its membership degree: The membership functions of water level include: The membership function of the low water level is A_L(x)=(Px) / P,x∈[0,P]; The membership function of the middle water level is A_M(x)=min(x / (P / 2),(Px) / (P / 2)),x∈[0,P]; The membership function of the high water level is A_H(x)=x / P,x∈[0,P]; The membership functions of traffic include: The membership function of low flow is A_S(y)=(Ny) / N,y∈[0,N]; The membership function of medium flow is A_M(y)=min(y / (N / 2),(Ny) / (N / 2)),y∈[0,N]; The membership function of high flow is A_L(y)=y / N,y∈[0,N]; S103, establishing fuzzy rules for water level, flow and drainage demand based on the binary comparison ranking method; specifically: If the water level is low and the flow is small, the drainage demand is low; If the water level is low and the flow is medium, the drainage demand is medium; If the water level is low and the flow is high, the drainage demand is high; If the water level is medium and the flow is small, the drainage demand is medium; If the water level is medium and the flow is medium, the drainage demand is medium; If the water level is moderate and the flow is high, the drainage demand is high; If the water level is high and the flow is low, the drainage demand is high; If the water level is medium and the flow is medium, the drainage demand is high; If the water level is medium and the flow is high, the drainage demand is high.

[0021] S104, obtaining output data through fuzzy reasoning and defuzzification; fuzzy reasoning calculates the membership of the output data according to the membership of the input data and the fuzzy rules; according to the fuzzy rules, the output data belongs to the membership of low, medium and high drainage demands.

[0022] Preferably, in step S104, according to the fuzzy rule, the specific method of outputting the membership of the data to low, medium and high drainage demands is: Low drainage demand membership: B_L=min(A_L(x),A_S(y)); Degree of subordination of drainage demand: B_M=max(min(A_L(x),A_M(y)),min(A_M(x),A_S(y)),min(A_M(x),A_M(y))); High drainage demand membership: B_H=max(min(A_L(x),A_L(y)),min(A_M(x),A_L(y)),min(A_H(x),A_S(y)),min(A_H(x),A_M(y)),min(A_H(x),A_L(y))); The comprehensive drainage demand is calculated by the centroid method: Y=(B_L*0+B_M*0.5+B_H*1) / (B_L+B_M+B_H); In the formula, x represents water level data and y represents flow data.

[0023] like Figure 2 As shown, preferably, in step S2, the specific method for constructing the start-stop rotation logic of the pump based on the membership of low, medium and high drainage demands is: S201, collecting historical drainage data of the drainage system and setting the priority of the drainage system: The drainage system includes the maintenance drainage system, the leakage drainage system and the corridor drainage system. The priorities of the drainage systems are set to F1, F2 and F3 according to the average value of the historical drainage volume; the drainage system with the smallest drainage volume is set to the highest priority; S202, setting the maximum drainage threshold of the drainage system with F1 priority to Q1, the maximum drainage threshold of the drainage system with F2 priority to Q2, and the maximum drainage threshold of the drainage system with F3 priority to Q3; S203, setting a drainage allocation strategy based on the comprehensive drainage demand Y: The drainage system with F1 priority is allocated for drainage first, the drainage demand exceeding Q1 is allocated to the drainage system with F2 priority, and the drainage demand exceeding Q2 is allocated to the drainage system with F3 priority; S204, collecting the working range and performance curve data of each pump in the maintenance drainage system, and determining the optimal drainage volume Q' of the pump from the performance curve; S205, numbering each pump, the numbered numbers forming an array Z; randomly sorting the numbered numbers in the array Z by a random algorithm to form a sorted pump sequence; S206, after receiving the allocated drainage demand, the drainage system starts the pumps in sequence according to the pump sequence in S205 to drain the water, and controls the power through the frequency converter of the pump. When the pump reaches the optimal drainage volume Q', it stops increasing the power and starts the next sequenced pump, and so on, until the drainage demand allocated by the drainage system can be matched; S207, when the drainage demand Y exceeds Q3, the controller synchronously increases the power of all the pumps started in the three drainage systems, evenly increases the power of all the pumps in use, so that the total drainage volume matches the comprehensive drainage demand Y; the drainage volume of each pump does not exceed its working range.

[0024] Furthermore, the random algorithm is implemented by using the Knuth shuffle algorithm to randomly arrange multiple numbered numbers; each time the drainage demand is redistributed, the start and stop sequence of the pumps is rearranged through the random algorithm.

[0025] Preferably, the maximum drainage thresholds Q1, Q2 and Q3 do not exceed the sum of the optimal drainage volumes Q' of all pumps in the corresponding drainage system.

[0026] like Figure 3 As shown, preferably, performing fault analysis on the operating state of the pump includes: The feedback signals of each drainage system are received through the computer system, including the real-time drainage data and power data of each pump; the fault diagnosis is performed on the collected data as follows: Data preprocessing: Clean, standardize and reduce noise of collected data to improve data quality and availability; Feature extraction: Use machine learning algorithms to extract features from preprocessed data and extract feature vectors that reflect the operating status of the equipment; Fault diagnosis: Use artificial intelligence algorithms such as classification, clustering, or regression to diagnose faults on feature vectors to determine whether the equipment has a fault and the type and extent of the fault.

[0027] Fault alarm: Based on the results of fault diagnosis, an alarm will be issued for the faulty equipment, including information on the fault location, time and cause, and corresponding maintenance suggestions will be provided.

[0028] Embodiment 2: This embodiment provides a specific fault diagnosis case: like Figure 4As shown in the figure, it is assumed that the drainage system of a hydropower station consists of a pump, a valve and two pressure sensors, where pressure sensor 1 measures the pressure P1 at the pump outlet and pressure sensor 2 measures the pressure P2 at the valve outlet. Under normal circumstances, P1 should be greater than P2, and the difference between the two should be within a certain range; if P1 is less than P2 or the difference between the two is out of range, it means that a fault has occurred.

[0029] Assuming that the normal data model is P1-P2=10±2 kPa, if it is detected that P1-P2 is not within this range, the data is judged to be abnormal and proceed to the next step.

[0030] By analyzing the characteristics of abnormal data, the type and location of faulty equipment can be determined.

[0031] For example, if P1-P2 < 8 kPa is detected, it means that the pressure at the pump outlet is too low, and the pump may be faulty; if P1-P2 > 12 kPa is detected, it means that the pressure at the valve outlet is too high, and the valve may be faulty; if P1 or P2 is detected to have a large deviation from historical data, it means that the corresponding pressure sensor is faulty. The corresponding fault diagnosis results and alarm signals are output according to different situations.

[0032] Take appropriate measures to correct the fault according to the fault diagnosis results; for example, if the pump fails, adjust the pump speed or replace the pump; if the valve fails, adjust the valve opening or replace the valve; if the pressure sensor fails, clean or replace the pressure sensor. And check whether the corrected pressure data returns to normal. If it returns to normal, end the diagnosis process; if it is still abnormal, repeat the above steps.

Claims

1. An intelligent joint control method applied to the drainage system of a hydropower station, characterized in that: The following steps are involved: S1, calculate the drainage demand of the hydropower station and divide the drainage demand of the hydropower station into low, medium and high drainage demands; According to the fuzzy rules, the membership degree and comprehensive drainage demand of low, medium and high drainage demand are output; S2, constructs the pump start-stop rotation logic based on the membership of low, medium and high drainage demands; S3, perform fault analysis on the operating status of the pump.

2. The intelligent joint control method for a drainage system of a hydropower station according to claim 1 is characterized in that: In step S1, the specific method for calculating the drainage demand of the hydropower station is: S101, assuming that the water level and flow rate ranges from [0, P] meters and [0, N] cubic meters per second, respectively, and the drainage demand ranges from [0, 1], where 0 indicates no drainage is required and 1 indicates the maximum drainage is required; Set low, medium and high thresholds for water level and flow respectively; divide water level and flow into several fuzzy subsets, including low, medium and high categories; S102, setting a membership function for each fuzzy subset to describe its membership degree: The membership functions of water level include: The membership function of the low water level is A_L(x)=(Px) / P,x∈[0,P]; The membership function of the middle water level is A_M(x)=min(x / (P / 2),(Px) / (P / 2)),x∈[0,P]; The membership function of the high water level is A_H(x)=x / P,x∈[0,P]; The membership functions of traffic include: The membership function of low flow is A_S(y)=(Ny) / N,y∈[0,N]; The membership function of medium flow is A_M(y)=min(y / (N / 2),(Ny) / (N / 2)),y∈[0,N]; The membership function of high flow is A_L(y)=y / N,y∈[0,N]; S103, establishing fuzzy rules for water level, flow and drainage demand based on binary comparison ranking method; S104, obtaining output data through fuzzy reasoning and defuzzification; fuzzy reasoning calculates the membership of the output data according to the membership of the input data and the fuzzy rules; according to the fuzzy rules, the output data belongs to the membership of low, medium and high drainage demands.

3. The intelligent joint control method for a drainage system of a hydropower station according to claim 2 is characterized in that: In step S104, according to the fuzzy rules, the specific method of outputting the degree of membership of the data belonging to low, medium and high drainage demands is as follows: Low drainage demand membership: B_L=min(A_L(x),A_S(y)); Degree of subordination of drainage demand: B_M=max(min(A_L(x),A_M(y)),min(A_M(x),A_S(y)),min(A_M(x),A_M(y))); High drainage demand membership: B_H=max(min(A_L(x),A_L(y)),min(A_M(x),A_L(y)),min(A_H(x),A_S(y)),min(A_H(x),A_M(y)),min(A_H(x),A_L(y))); The comprehensive drainage demand is calculated by the centroid method: Y=(B_L*0+B_M*0.5+B_H*1) / (B_L+B_M+B_H); In the formula, x represents water level data and y represents flow data.

4. The intelligent joint control method for a drainage system of a hydropower station according to claim 3 is characterized in that: In step S2, the specific method of constructing the pump start-stop rotation logic based on the membership of low, medium and high drainage demands is: S201, collecting historical drainage data of the drainage system and setting the priority of the drainage system: The drainage system includes the maintenance drainage system, the leakage drainage system and the corridor drainage system. The priorities of the drainage systems are set to F1, F2 and F3 according to the average value of the historical drainage volume; the drainage system with the smallest drainage volume is set to the highest priority; S202, setting the maximum drainage threshold of the drainage system with F1 priority to Q1, the maximum drainage threshold of the drainage system with F2 priority to Q2, and the maximum drainage threshold of the drainage system with F3 priority to Q3; S203, setting a drainage allocation strategy based on the comprehensive drainage demand Y: The drainage system with F1 priority is allocated for drainage first, the drainage demand exceeding Q1 is allocated to the drainage system with F2 priority, and the drainage demand exceeding Q2 is allocated to the drainage system with F3 priority; S204, collecting the working range and performance curve data of each pump in the maintenance drainage system, and determining the optimal drainage volume Q' of the pump from the performance curve; S205, numbering each pump, the numbered numbers forming an array Z; randomly sorting the numbered numbers in the array Z by a random algorithm to form a sorted pump sequence; S206, after receiving the allocated drainage demand, the drainage system starts the pumps in sequence according to the pump sequence in S205 to drain the water, and controls the power through the frequency converter of the pump. When the pump reaches the optimal drainage volume Q', it stops increasing the power and starts the next sequenced pump, and so on, until the drainage demand allocated by the drainage system can be matched; S207, when the drainage demand Y exceeds Q3, the controller synchronously increases the power of all the pumps started in the three drainage systems, evenly increases the power of all the pumps in use, so that the total drainage volume matches the comprehensive drainage demand Y; the drainage volume of each pump does not exceed its working range.

5. The intelligent joint control method for a drainage system of a hydropower station according to claim 4 is characterized in that: The maximum discharge thresholds Q1, Q2 and Q3 do not exceed the sum of the optimal discharge capacities Q' of all pumps in the corresponding drainage system.

6. The intelligent joint control method for a drainage system of a hydropower station according to claim 5, characterized in that: Fault analysis of the pump's operating status includes: The feedback signals of each drainage system are received through the computer system, including the real-time drainage data and power data of each pump; the fault diagnosis is performed on the collected data as follows: Data preprocessing: Clean, standardize and reduce noise of collected data to improve data quality and availability; Feature extraction: Use machine learning algorithms to extract features from preprocessed data and extract feature vectors that reflect the operating status of the equipment; Fault diagnosis: Use artificial intelligence algorithms such as classification, clustering, or regression to diagnose faults on feature vectors to determine whether the equipment has a fault and the type and extent of the fault. Fault alarm: Based on the results of fault diagnosis, an alarm will be issued for the faulty equipment, including information on the fault location, time and cause, and corresponding maintenance suggestions will be provided.

Images