A load distribution method for a hydropower plant-level automatic power generation control system
By introducing a model predictive control unit in the automatic power generation control system of a hydropower plant, comprehensively considering multiple operating indicators of the unit and optimizing the load distribution plan, the unit vibration and efficiency problems caused by unreasonable load distribution in the existing technology are solved, and Improved load regulation response speed.
Patent Information
- Application Number
- CN202210362452.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-04-07
- Publication Date
- 2025-02-07
- Estimated Expiration
- 2042-04-07
AI Technical Summary
The existing automatic power generation control system of hydropower plants fails to comprehensively consider the working head, vibration swing and other factors of the unit during load distribution, resulting in the unit's vibration and efficiency often ignoring one by one, and the large amount of calculation results in slow load regulation response.
The model prediction control unit is introduced in the automatic power generation control system of the hydropower plant. Combined with linear planning and control issues, comprehensively considering the working head of the unit, vibration swing, active load, efficiency indicators, the number of times of passing through the vibration zone and the cumulative adjustment times of the unit, and formulate the Load distribution plan.
By taking into account multiple factors, the load distribution plan is optimized, the unit vibration is reduced, the operation stability is improved, and the load regulation response time is significantly shortened.
Smart Images

Figure CN114759611B_ABST
Abstract
Description
Technical Field
[0001] The invention belongs to the field of plant-level load dispatching of hydropower plants, and mainly relates to a load distribution method of a plant-level automatic power generation control system of a hydropower plant. Background Art
[0002] As the proportion of power generation load of large hydroelectric generating units in the total load of the power grid continues to increase, the power grid has higher and higher requirements for the safety, stability, and load regulation performance of large hydroelectric generating units. Large hydroelectric generating units should be able to quickly respond to changes in power grid load instructions, while also ensuring the safe, stable, and efficient operation of the generating units themselves.
[0003] How to reasonably distribute the total load of the whole plant to each unit in the plant, and how to allocate the load increment or load reduction to each unit during the load adjustment process is a complex decision-making process, which is crucial to the safe, stable and efficient operation of the unit. Unreasonable load distribution and adjustment methods will lead to low power generation efficiency of the unit, increased vibration of the unit, increased maintenance frequency, shortened equipment service life, and even major safety accidents.
[0004] Large hydropower plants usually use plant-level automatic power generation control systems to distribute the total plant load issued by the dispatching master station to each unit in the plant. When load regulation is carried out, the plant-level automatic power generation control system is also used to determine whether one or several units in the plant will bear the load increase or load decrease during the regulation process.
[0005] The load distribution of the existing hydropower plant-level automatic power generation control system is generally distributed according to the following principles:
[0006] 1) Average distribution principle: The total load of the whole plant is evenly distributed to each unit in the plant; during the load adjustment process, the increase or decrease of the total load of the whole plant is also evenly distributed to each unit. The advantage of this method is that the load distribution and execution operation are simple and fast. The disadvantages are: ① The performance differences between different units are not taken into account; ② All units participate in each load adjustment process, and the cumulative adjustment times of the units are large, which affects the service life of the equipment.
[0007] 2) Equal incremental rate principle: the total load of the whole plant is distributed in descending order according to the efficiency of the units, and the load is preferentially distributed to the units with small incremental rate (i.e. high efficiency); during the load adjustment process, when the power plant increases the load, the units with small incremental rate are allowed to increase the load first; when the power plant reduces the load, the units with large incremental rate are allowed to reduce the load first. That is, in general, the units with high efficiency are allowed to maintain a higher load rate and a longer operating time.
[0008] 3) Small load distribution strategy: During the load regulation process, when the power plant increases the load, priority is given to units with a small load rate to increase the load; when the power plant reduces the load, priority is given to units with a large load rate to reduce the load. The advantage of this is that it can effectively reduce the frequency of unit regulation and avoid the unit's regulation dead zone.
[0009] There are many factors to consider in the process of formulating a load distribution plan. The goals behind each factor are also different, and sometimes they even conflict with each other and cannot be unified. The determination of the load distribution plan often requires a trade-off between different considerations, focusing on the big and letting go of the small, and giving priority to solving the main contradictions. In the process of load distribution and load regulation throughout the plant, it is necessary not only to ensure the efficient operation of the unit as much as possible, but also to reduce the vibration swing of the unit as much as possible, reduce the number of times the unit passes through the vibration zone, and reduce the total duration of the unit's operation in the vibration zone; in addition, the cumulative number of adjustments of each unit should be kept as equal as possible, and it is not possible to simply let the unit with good performance undertake most of the adjustment tasks; under different load rate conditions, the most suitable load distribution strategy should be flexibly selected, such as adopting a high-efficiency priority distribution strategy during the flood season with high load rate, and adopting a safety and stability priority load distribution strategy during the dry season with low load rate.
[0010] When the existing hydropower plant-level automatic power generation control system distributes load, the main factors considered are the power and efficiency of the unit, without taking into account the working head, vibration swing, number of times crossing the vibration zone, cumulative number of adjustments and other factors in the load distribution and adjustment process. In the actual operation process, it is often the case that one thing is neglected while another is neglected. For example, the efficiency values of all units in the plant are good, but the unit vibration problem is serious, and the vibration zone is frequently crossed during the adjustment process; or among multiple units, one unit with good comprehensive performance has a much larger number of adjustments than other units, resulting in its maintenance cycle being significantly shorter than other units.
[0011] The existing calculation procedures of the hydropower plant-level automatic power generation control system mainly adopt two algorithms: the exhaustive method and the equation solution method.
[0012] The exhaustive method is to arrange and combine all the load distribution conditions of the unit to form a data list of hundreds of millions of records; when load adjustment is to be performed, hundreds of thousands of records that meet the requirements are retrieved from the data list according to the adjustment target, and then each record is scored according to certain rules. After sorting according to the scores, the one with the highest score is selected as the load distribution plan.
[0013] The equation-solving method is to sort out a set of equations with constraints based on the factors to be considered in load distribution, and then find the optimal solution through matrix calculation and iterative convergence.
[0014] Both algorithms require large amounts of computation and take a long time to calculate, resulting in slow response of the unit to grid load regulation instructions. Summary of the invention
[0015] The purpose of the present invention is to provide a load distribution method for a hydropower plant-level automatic power generation control system to solve the problems in the prior art that the load distribution strategy is imperfect, resulting in the vibration and efficiency of the unit often being compromised; and the program has a large amount of calculation and is time-consuming, resulting in slow load regulation response.
[0016] The technical solution of the present invention is: to introduce a model prediction control unit into the plant-level automatic power generation control system of the hydropower plant, the model prediction control unit combines linear programming and control problems, and comprehensively considers the working water head, vibration swing, active load, efficiency index, number of times of crossing the vibration zone, and cumulative adjustment number of the unit in the process of formulating the load distribution plan of the plant-level automatic power generation control system of the hydropower plant and selecting the best plan.
[0017] The above model prediction control unit receives the real-time operating parameters of the unit sent back by the data acquisition system as the initial conditions for load regulation. The real-time operating parameters of the unit include the working water head ( )、vibration swing(S)、active load(P)、efficiency index( )、Gate opening( ). The relationship between the various operating parameters of the unit can be organized into a set of equations with constraints: ,The constraints include the working head range, active power range, and vibration swing range. According to the constraints, different operating performance areas can be divided on the operating area diagram of each unit, including small load vibration area, unrecommended operation area, prohibited operation area, and high load operation stable area.
[0018] The steps for the above model prediction control unit to formulate a load distribution plan include:
[0019] 1) Receive the load instructions and real-time operating parameters of the unit issued by the general dispatcher and determine the load adjustment amount;
[0020] 2) Solve the problem of including the unit working head , vibration swing S, active load P, efficiency index , gate opening ( )The unit operation state equation of these five variables , get all possible load distribution schemes;
[0021] 3) Select the best load distribution scheme among all possible load distribution schemes as the final load distribution scheme to be executed;
[0022] 4) Verify the adjusted state of the unit. Once confirmed to be correct, select the best load distribution plan and send it to the actuator for execution.
[0023] The steps of selecting the best load distribution scheme by the above model prediction control unit include:
[0024] 1) Taking into account the constraints of the unit's working water head, vibration swing, active load and efficiency index, different operating areas are divided on the operating area diagram of each unit, including prohibited operating area, small load vibration area, not recommended operating area, and high load stable operating area;
[0025] 2) When selecting the best load distribution scheme, first exclude the data falling into the prohibited operation area, and then search for available load distribution schemes from the high-load stable operation area, the unrecommended operation area, and the low-load vibration area in turn according to the principle of high efficiency priority or the principle of small vibration swing priority; the principle of high efficiency priority and the principle of small vibration swing priority can be switched. The principle of high efficiency priority is adopted during the high-load period of the unit operation, and the principle of small vibration swing priority is adopted during the low-load period of the unit operation;
[0026] 3) When selecting the best load distribution plan, ensure that the number of times each unit passes through the vibration zone and the cumulative number of adjustments are basically the same, to prevent high-performance units with high efficiency and small vibration swing from taking on too many adjustment tasks and shortening their maintenance cycle.
[0027] The above-mentioned model prediction control unit can collect the real-time operating parameters of the unit, and score the performance of the unit from the two dimensions of efficiency and vibration swing. The operating area can be re-divided on the operating area map of each unit according to the score; when selecting the best load distribution plan, the area with low performance score is first excluded, and the area with high performance score is given priority. This can narrow the scope of data screening, greatly reduce the amount of calculation, and improve the response speed to load adjustment instructions. As the unit's operating time accumulates, the performance of the unit itself will change, and the division of the operating areas of each unit will also change accordingly; for example, the original non-recommended operating area may be classified as a prohibited operating area due to increased vibration, and part of the original high-load stable operating area may be classified as a non-recommended operating area.
[0028] The above-mentioned model prediction control unit is provided with an on / off switch button, which can deactivate the model prediction control unit. After the model prediction control unit is deactivated, the normal operation of the hydropower plant-level automatic power generation control system is not affected. BRIEF DESCRIPTION OF THE DRAWINGS
[0029] Figure 1 This is a control block diagram of the present invention when applied to the Tianshengqiao Level 1 Power Station
[0030] Figure 2 This is the typical operating area diagram of the four units of the Tianshengqiao Level 1 Power Station
[0031] Figure 3The operating area diagram of the 2# and 3# units is re-divided according to the actual conditions of the units.
[0032] Figure 4 The operating area diagram of the 1# and 4# units is re-divided according to the actual conditions of the units. DETAILED DESCRIPTION
[0033] The present invention will be further described below in conjunction with the accompanying drawings and specific implementation methods, but the present invention is not limited to the following embodiments.
[0034] The total installed capacity of Tianshengqiao Level 1 Power Station is 1,200MW, with four units of the same model and a single capacity of 300MW. Figure 2 This is a typical operating area diagram of the four units of the Tianshengqiao Level 1 Power Station. The load distribution method of the original plant-level automatic power generation control system is to solve the unit operating state equation , determine the load borne by each unit and the increase / decrease in load borne during the adjustment process. The following problems occurred in actual operation: ① Each unit operated in the vibration zone for a long time, and the vibration caused frequent cracks in the rotor blades of units 1# and 4#, and even local block loss; ② The load adjustment of each unit was too frequent, and the cumulative number of adjustments was large, especially the 2# unit, because the rotor was new and in good working condition, and the number of adjustments was significantly more than other units; ③ Frequently crossed the vibration zone during the adjustment process; ④ The unit responded slowly to the load command.
[0035] The control block diagram of the present invention when applied to the Tianshengqiao primary power station is as follows: Figure 1 As shown in the figure, a model prediction control unit is introduced into the plant-level automatic power generation control system. The model prediction control unit in the figure receives the real-time operating status parameter working head of the unit , vibration swing S, active load P, efficiency index , gate opening ( ) as the initial condition for load regulation, the model prediction control unit solves the unit operation state equation Determine the load distribution plan. Compared with the original load distribution method, two constraints, working water head and vibration swing, are added to the unit operation state equation. When selecting the best load distribution plan, both efficiency and vibration conditions are considered, and the unit's operating state is prevented from falling into the vibration zone while pursuing high efficiency. In actual use, the operating condition when the vibration swing exceeds 0.3mm is set as a prohibited operating condition, and the operating condition when the vibration swing is 0.15~0.3mm is set as a not recommended operating condition.
[0036] As the unit is used for a long time, its working characteristics will also change. The model prediction control unit collects the real-time operating parameters of the unit, scores the performance of the unit from two dimensions: efficiency index and vibration swing, and re-divides the operating area on the operating area map of each unit according to the score. Figure 3 This is the operation area diagram of Units 2# and 3# that was re-divided according to the actual conditions of the units. Figure 4 The operation area diagram of the 1# and 4# units is re-divided according to the actual conditions of the units. Figure 2 Compared with the typical operation area diagram of the unit, the division of the new operation area diagram is significantly different. The operation area diagram of the 2# and 3# units has added a non-recommended operation area, the range of the small load vibration area has been expanded, and the optional operation area is divided into the optimal area and the suboptimal area; the operation area diagram of the 1# and 4# units has added a prohibited operation area and a non-recommended operation area, and the optional operation area is divided into the optimal area and the suboptimal area.
[0037] contrast Figure 2 and Figure 3 / Figure 4 From the operating area diagram, we can see that Figure 2 A considerable part of the area that was originally considered to be a high-load stable operation area is actually in a very bad operating state and can be classified as a non-recommended operation area or even a prohibited operation area. This is also an important reason why the unit often vibrates violently when the original load distribution method is used.
[0038] When the model prediction control unit selects the best load distribution plan, it first avoids the prohibited operation area and the not recommended operation area, while reducing the number of crossing the prohibited operation area and the not recommended operation area, and considers selecting the unit's operating parameters within the optimal area.
[0039] By dividing the prohibited operation area and the unrecommended operation area, the load distribution avoids the operation area where the unit vibrates violently, and the average maximum vibration speed of the unit is reduced from 50cm / s to no more than 10cm / s, which greatly improves the operation stability of the unit. At the same time, due to the narrowing of the selection range of the allocation plan, the amount of calculation is greatly reduced, and the average response speed to the load adjustment command is also shortened from 10~15ms to 6~8ms, with a speed increase of nearly 50%.
Claims
1. A load distribution method for a hydropower plant-level automatic power generation control system, characterized in that: It includes a model prediction control unit, which can comprehensively consider the unit's working water head, vibration swing, active load, efficiency index, number of times of crossing the vibration zone, and cumulative number of adjustments to formulate and select the best load distribution plan; The steps of the model prediction control unit formulating the load distribution plan include: 1) Receive the load instructions and real-time operating parameters of the unit issued by the general dispatcher and determine the load adjustment amount; 2) Solve the problem of including the unit working head , vibration swing S, active load P, efficiency index , gate opening ( )The unit operation state equation of these five variables , get all possible load distribution schemes; 3) Select the best load distribution scheme among all possible load distribution schemes as the final load distribution scheme to be executed; 4) Verify the adjusted parameters of the unit. After confirming that they are correct, select the best load distribution plan and send it to the actuator for execution; The step of selecting the best load distribution scheme by the model prediction control unit includes: 1) Taking into account the constraints of the unit's working water head, vibration swing, active load and efficiency index, different operating areas are divided on the operating area diagram of each unit, including prohibited operating area, small load vibration area, not recommended operating area, and high load stable operating area; 2) When selecting the best load distribution scheme, first exclude the data falling into the prohibited operation area, and at the same time reduce the number of crossing the prohibited operation area and the not recommended operation area, and select the best operating parameters of the unit in the optimal area; then, according to the principle of high efficiency priority or the principle of small vibration swing priority, select the best load distribution scheme from the high-load stable operation area, the not recommended operation area, and the small load vibration area in turn; the principle of high efficiency priority and the principle of small vibration swing priority can be switched. The principle of high efficiency priority is adopted during the high-load period of the unit operation, and the principle of small vibration swing priority is adopted during the low-load period of the unit operation. The operating condition when the vibration swing exceeds 0.3mm is set as the prohibited operation condition, and the operating condition when the vibration swing is 0.15~0.3mm is set as the not recommended operation condition. By dividing the prohibited operation area and the not recommended operation area, the load distribution avoids the operation area where the unit vibrates violently; 3) When selecting the best load distribution plan, ensure that the number of times each unit passes through the vibration zone and the cumulative number of adjustments are basically the same, to prevent high-performance units with high efficiency and small vibration swing from taking on too many adjustment tasks; The model prediction control unit can collect the real-time operating parameters of the unit, and score the performance of the unit from the two dimensions of efficiency index and vibration swing, and re-divide the operating area on the operating area map of each unit according to the score; when selecting the best load distribution plan, first exclude the area with low performance score, give priority to the area with high performance score, and narrow the scope of data screening; The model prediction control unit is provided with an on / off switch button, which can deactivate the model prediction control unit. Deactivation of the model prediction control unit does not affect the normal operation of the plant-level automatic power generation control system.
Citation Information
Patent Citations
Vibration-area-crossing method of automatic gain control (AGC) system for different water heads in hydroelectric plant
CN103036232A
System for distributing loads among plants of cascade hydropower stations in real time and implementation method of system
CN103971198A