A method and system for coordinated control of cascade hydropower stations
By constructing a hydraulic coupling transfer function and dynamic reserve capacity allocation, combined with unit aging correction and flow change rate, the problem of inaccurate assessment of hydraulic coupling relationship in cascade hydropower stations was solved, achieving safe and reliable regulation and rapid response, and improving the operational stability and economy of cascade hydropower stations.
Patent Information
- Application Number
- CN202611093610.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2026-07-22
- Publication Date
- 2026-11-17
- Estimated Expiration
- 2046-07-22
AI Technical Summary
Existing cascade hydropower station control technologies fail to effectively quantify and analyze hydraulic coupling relationships, resulting in inaccurate reserve capacity assessments, an inability to respond to head changes in real time, a lack of safe and feasible domain construction and feedforward compensation mechanisms, and potential equipment safety hazards during regulation.
By constructing a transfer function of the impact of upstream power plant outflow on downstream power plant head, dynamic positive and negative reserve capacity is allocated in real time. Combined with unit aging correction and flow change rate, a safe and feasible domain is constructed, and a rolling time-domain optimization algorithm is used for feedforward compensation to achieve multivariate adaptive correction.
It improves the regulation safety and reliability of cascade hydropower stations, can proactively prevent damage to equipment caused by hydraulic transients, and enhances the speed and economy of responding to grid dispatch instructions.
Smart Images

Figure CN122600342B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of automatic control technology for hydropower stations, and more specifically, to a coordinated control method and system for cascade hydropower stations. Background Technology
[0002] A cascade hydropower station is a complex hydropower system consisting of multiple hydropower stations located at different altitudes on the same river. The stations are connected by water flow, forming a close hydraulic coupling relationship. The outflow from the upstream station directly affects the inflow and head of the downstream station. With the increasing demands of power systems for clean energy regulation capabilities, cascade hydropower stations need to simultaneously undertake the dual tasks of economic dispatch control (EDC) and automatic generation control (AGC). The former focuses on maximizing power generation efficiency while meeting various constraints, while the latter requires rapid response to changes in grid load to maintain system frequency stability.
[0003] Existing cascade hydropower station control technologies primarily employ a layered, decoupled control architecture, treating EDC and AGC as independent control components. At the economic dispatch level, they typically rely on deterministic inflow forecasts and fixed head assumptions, using dynamic programming or constant incremental rate methods to formulate day-ahead or intraday power generation plans. This approach fails to adequately consider the dynamic changes in real-time hydraulic conditions and the time-varying nature of unit operating states. At the automatic generation control level, simple proportional allocation or fixed coefficient allocation methods based on installed capacity are commonly used to distribute grid dispatch commands to each power station. This lacks real-time quantitative analysis of upstream and downstream hydraulic coupling effects, resulting in the inability to promptly detect and compensate for downstream head changes caused by upstream power station regulation. Furthermore, existing technologies rely heavily on manual experience to set fixed thresholds for handling issues such as unit vibration zone avoidance and hydraulic transient constraints in the water conveyance system, making them ill-suited to complex and variable operating conditions.
[0004] In recent years, some studies have attempted to introduce advanced methods such as model predictive control and adaptive control, but the following limitations still exist: First, the establishment of hydraulic coupling models is mostly based on steady-state assumptions, failing to effectively characterize the dynamic transmission characteristics of outflow and downstream head changes under unsteady flow conditions; second, the calculation of reserve capacity usually uses fixed design values, without considering the dynamic impact of head changes, unit aging, ecological constraints, and other factors; third, the lack of real-time safety monitoring of the flow rate and head change rate in the water conveyance channel during the regulation process easily leads to equipment safety problems caused by hydraulic transients; fourth, the coordination mechanism between feedforward control and feedback correction is imperfect, and the degree of integration between predictive information and real-time operating data is low.
[0005] In implementing the embodiments of the present invention, the prior art has at least the following problems or defects: the prior art fails to establish a dynamic transfer function of the impact of upstream power plant outflow on downstream power plant head, and cannot quantitatively analyze the real-time correction requirements of hydraulic coupling relationship on reserve capacity; the prior art uses fixed upper and lower limits of output to calculate reserve capacity, without considering the dynamic impact of head change rate, head change acceleration and unit aging state on regulation capacity; the prior art's regulation deviation allocation method does not distinguish the differences in hydraulic coupling effect under positive and negative regulation directions, resulting in inaccurate reserve capacity assessment of downstream power plants during upstream regulation; the prior art lacks a safe feasible domain and graded safety restriction measures based on flow change rate and head change rate, and cannot implement preventive control before the accumulation of hydraulic transient risks; the prior art's feedforward compensation mechanism fails to effectively integrate load forecast and inflow forecast information, and the adjustment of reserve capacity allocation ratio lags behind actual operating condition changes; the prior art's correction mechanism is mostly a single parameter adjustment, making it difficult to achieve multi-variable collaborative optimization of allocation method, unit sequencing and safety boundary. Summary of the Invention
[0006] This invention provides a method and system for coordinated control of cascade hydropower stations.
[0007] In a first aspect of the present invention, a coordinated control method for a cascade hydropower station is provided, comprising: S1: Collect real-time operating status parameters, hydraulic environment parameters, and total active power commands issued by the dispatching system for each power station in the cascade hydropower station. Calculate the current total output, regulation deviation, dynamic positive reserve capacity, and dynamic negative reserve capacity based on the collected data. S2: Based on the sign of the adjustment deviation, the dynamic positive reserve capacity and the dynamic negative reserve capacity of each power station, and combined with the coupling relationship between the outflow of the upstream power station and the head of the downstream power station, the adjustment deviation is allocated in multiple levels to generate the allocation adjustment amount of each power station. S3: Determine the unit regulation sequence based on the comprehensive regulation capacity index of each power station's internal units, and generate the unit regulation sequence; S4: According to the unit adjustment sequence, the unit is controlled to perform output adjustment in sequence by step approximation method, and the output limit, vibration zone boundary and head change feedback of each unit are monitored in real time during the adjustment process until the distribution adjustment of each power station is completed. S5: During the output regulation process, a safe and feasible domain is constructed based on the flow rate change rate and head change rate of the water conveyance channel, and dynamic safety limits are imposed on the regulation rate and regulation increment according to real-time hydraulic parameters. S6: Based on the predicted load change trend and inflow flow change trend, feedforward compensation correction is performed for the regulation cycle that has not yet occurred, and the reserve capacity allocation ratio of each power station is adjusted in advance. S7: Based on the deviation between the actual total output and the dispatch command after the end of each adjustment cycle, as well as the deviation between the actual hydraulic parameters and the predicted hydraulic parameters, multivariate corrections are made to the adjustment deviation allocation method, the unit adjustment sequence, and the safe and feasible domain boundary.
[0008] In a second aspect of the present invention, a coordinated control system for a cascade hydropower station is provided, comprising: The data acquisition module is used to collect real-time operating status parameters, hydraulic environment parameters, and total active power commands issued by the dispatching system for each power station in the cascade hydropower station. The dynamic reserve calculation module is used to correct the upper and lower limits of output based on the current head, and to calculate the dynamic positive reserve capacity and dynamic negative reserve capacity based on the corrected upper and lower limits of output and the current output. The hydraulic coupling distribution module is used to distribute the adjustment deviation in multiple levels according to the sign of the adjustment deviation, the dynamic reserve capacity of each power station, and the transfer function of the influence of the outflow of the upstream power station on the head of the downstream power station, and to generate the distribution adjustment amount of each power station. The unit sequencing module is used to calculate the comprehensive regulation capacity index based on the unit's remaining regulation capacity, efficiency characteristic coefficient, vibration zone penalty factor, and fatigue correction coefficient, and to generate the unit regulation sequence according to the comprehensive regulation capacity index. The adjustment execution module is used to control the unit to perform output adjustment within the adjustable output range according to the unit adjustment sequence and using a step-approximation method. The adjustable output range is dynamically determined based on the dynamic output upper and lower limits and the vibration range boundary. The safety constraint module is used to construct a safe and feasible region based on the rate of change of flow and the rate of change of head in the water conveyance channel, and to implement graded safety constraint measures when the regulation process exceeds the safe and feasible region, including reducing the regulation rate, limiting the regulation increment, suspending regulation, and reverse back-off regulation; The feedforward prediction module is used to generate feedforward compensation based on load change trends and inflow flow trends using a rolling time-domain optimization algorithm, and to adjust the reserve capacity allocation ratio of each power station in advance. The correction module is used to perform multivariate adaptive correction of the allocation correction coefficient, sorting weighting coefficient, and safe feasible domain boundary scaling coefficient based on the periodic output deviation and hydraulic prediction deviation, using the gradient descent method.
[0009] The embodiments of the present invention have at least the following beneficial effects: 1. By constructing a transfer function for the impact of upstream power station outflow on downstream power station head, real-time quantitative analysis of hydraulic coupling relationship is achieved. During the process of adjusting deviation allocation, the dynamic positive or negative reserve capacity of downstream power station can be dynamically corrected according to the changes in upstream output. This effectively solves the problem of inaccurate assessment of downstream power station reserve capacity caused by neglecting the hydraulic coupling effect in existing technologies. It avoids the situation where the actual adjustable capacity of downstream power station decreases due to the rise in head when upstream power station increases output, or the risk of over-output increases due to the drop in head when upstream power station decreases output. This improves the safety and reliability of joint regulation of cascade hydropower stations.
[0010] 2. A dynamic correction mechanism based on the current head, head change rate, and head change acceleration is adopted, combined with unit aging correction factors and regulation reliability coefficients, to achieve refined calculation of dynamic positive and negative reserve capacity. At the same time, by constructing a real-time safe and feasible domain with upper limits for flow rate change and head change rate as boundaries, and setting early warning intervals to implement graded safety restriction measures, the problem that fixed upper and lower limits of output and fixed safety thresholds in existing technologies cannot adapt to complex hydraulic transient conditions is effectively solved. It can proactively reduce the regulation rate or suspend regulation before hydraulic parameters approach the safety boundary, prevent hydraulic transients in the water conveyance system from damaging the equipment, and ensure the long-term safe and stable operation of the cascade hydropower station.
[0011] 3. By integrating load change trend prediction and inflow flow change trend prediction through a rolling time-domain optimization algorithm to generate feedforward compensation, the reserve capacity allocation ratio of each power station is adjusted in advance. At the same time, based on the periodic output deviation and hydraulic prediction deviation, the gradient descent method is used to perform multivariate adaptive correction of the allocation correction coefficient, sorting weighting coefficient, and safe feasible domain boundary scaling coefficient. This effectively solves the problems of insufficient integration of feedforward information and real-time data and the single and lagging correction mechanism in the existing technology. It realizes deep synergy between predictive control and feedback control, enabling cascade hydropower stations to respond to grid dispatch instructions with both speed and economy, and significantly improves the dynamic quality and adaptive capability of joint EDC and AGC coordinated control. Attached Figure Description
[0012] The above and other objects, features, and advantages of exemplary embodiments of the present invention will become readily apparent from the following detailed description taken in conjunction with the accompanying drawings. Several embodiments of the invention are illustrated in the drawings by way of example and not limitation, wherein: Figure 1 This is a flowchart illustrating a coordinated control method for cascade hydropower stations provided in an embodiment of the present invention. Figure 2 This is a schematic diagram of the structure of a cascade hydropower station coordination control system provided in an embodiment of the present invention. Detailed Implementation
[0013] The principles and spirit of the invention will now be described with reference to several exemplary embodiments. It should be understood that these embodiments are provided merely to enable those skilled in the art to better understand and implement the invention, and are not intended to limit the scope of the invention in any way. Rather, these embodiments are provided to make the invention more thorough and complete, and to fully convey the scope of the invention to those skilled in the art.
[0014] The following is for reference. Figure 1 , Figure 1 This is a flowchart illustrating a coordinated control method for cascade hydropower stations provided in an embodiment of the present invention.
[0015] like Figure 1 As shown, a coordinated control method for cascade hydropower stations includes: S1: Collect real-time operating status parameters, hydraulic environment parameters, and total active power commands issued by the dispatching system for each power station in the cascade hydropower station. Calculate the current total output, regulation deviation, dynamic positive reserve capacity, and dynamic negative reserve capacity based on the collected data.
[0016] A cascade hydropower station refers to multiple hydropower stations built sequentially on the same river, with the outflow from the upstream station flowing to the downstream station, forming a hydraulic coupling relationship. Head refers to the height difference between the upstream and downstream water levels of a hydropower station, measured in meters, and is a key parameter affecting the output of the turbine generator units. Output refers to the active power generated by the turbine generator units, measured in megawatts (MW). Regulation deviation refers to the difference between the total active power command issued by the dispatching system and the current total output of the cascade hydropower stations. Reserve capacity refers to the difference between the current output of the power station and its upper or lower limit; positive reserve capacity indicates the capacity that can be increased, while negative reserve capacity indicates the capacity that can be reduced.
[0017] The data acquisition module obtains real-time operating status parameters of each power station, including current active power output, current head, and water flow in the water conveyance channel. Simultaneously, it receives total active power commands from the power grid dispatching system. The data acquisition module consists of a programmable logic controller (PLC) or a remote terminal unit (RTU), communicates with the dispatching center using the IEC 60870-5-104 protocol, and has a sampling period of 1 second.
[0018] Calculations based on collected data: Current total output ,in For the first The current active power output of each power station Total number of power stations; adjustment deviation ,in For scheduling instructions; dynamic positive reserve capacity and dynamic negative reserve capacity The calculation method will be explained in subsequent steps. The above calculations provide the basic data for subsequent power allocation.
[0019] S11: Collect real-time operating status parameters of each power station in the cascade hydropower station. These real-time operating status parameters include the current active power output, current head, current water conveyance channel flow, operating status identifier of each unit, and historical output sequence of each unit.
[0020] Active power output is measured via a power transmitter with an accuracy of 0.2%. The current head is obtained by measuring the difference between upstream and downstream water level gauges, which can be ultrasonic or pressure gauges. The current flow rate in the water conveyance channel is measured using an ultrasonic or electromagnetic flow meter. The operating status identifier is an enumerated variable, with values for operating, shut down, under maintenance, or standby, provided by the unit control system, such as the governor PLC. Historical output sequences are stored in a real-time database, saving output values for the past 60 regulation cycles, each regulation cycle being 1 second. This step collects the above parameters from the distributed control systems (DCS) of each power station via the OPC UA protocol.
[0021] S12: Collect the total active power command issued by the scheduling system, and simultaneously collect the timestamp and valid identifier of the total active power command.
[0022] The dispatching system issues commands through the power dispatching data network, and the message format conforms to the IEC 60870-5-104 protocol. The total active power command is a 16-bit integer value, with the unit being 0.1 megawatts. The timestamp is UTC time, accurate to milliseconds, synchronized with the GPS clock source at the dispatching center. The command validity identifier is a 1-byte value, with a value of 0xAA indicating validity and other values indicating invalidity. The communication server parses these three fields after receiving the message.
[0023] S13: Perform data preprocessing on the current active power output of each power station. This data preprocessing includes removing outliers that are significantly beyond the equipment's range, interpolating and filling in missing data, and performing first-order low-pass filtering on the data after removal and filling to generate the filtered active power output of each power station.
[0024] Assume the power transmitter range is ,in This is the minimum measurable value (i.e., the lower limit of the range) of the power transmitter. This represents the maximum measurable value (i.e., the upper limit of the measurement range) of the power transmitter. If the acquired value... Values outside this range are considered outliers and removed. For missing data due to communication interruptions or other reasons, linear interpolation is used to fill in the missing data: Let the missing time be... The effective values before and after are respectively and The corresponding times are respectively and Then fill value .
[0025] The first-order low-pass filtering algorithm is ,in This is the current sampled value (after removing padding). The output of the previous filtering, the filter coefficients This filter can suppress high-frequency noise and smooth the output data. This step can be implemented in the PLC using a circular buffer, or by writing a filter function in C++ in the host computer software.
[0026] S14: Perform a cumulative summation operation on the filtered active power output of each power station, and use the cumulative result as the current total output.
[0027] Current total output ,in For the first The active power output of a power station after passing through the S13 filter. This calculation is performed by the adder in the allocation calculation module. In embedded systems, it can be implemented using an accumulation instruction, while in PLCs, it can be implemented using ADD blocks in series.
[0028] S15: Perform a validity check on the total active power command. This validity check includes checking whether the command validity identifier is valid and whether the difference between the timestamp and the current system time is less than the preset maximum allowable delay. If the check passes, the total active power command is used as a valid command in subsequent calculations. If the check fails, the valid command from the previous cycle is used and an invalid command alarm is generated.
[0029] The maximum allowed delay is set to 10 seconds. (If the command is valid...) And the current system time Subtract timestamp If the time is less than 10 seconds, the instruction is valid, and the valid instruction cache is updated. Otherwise, keep The alarm remains unchanged, and an alarm is triggered via a pop-up window on the human-machine interface if the dispatch command has timed out or is invalid. At the same time, the alarm event is written to the event log file for maintenance personnel to review.
[0030] S16: Calculate the difference between the effective command and the current total output, use this difference as the original adjustment deviation, and perform dead-zone correction on the original adjustment deviation based on the preset adjustment dead zone to generate the adjustment deviation for subsequent allocation. The specific method of dead-zone correction is as follows: when the absolute value of the original adjustment deviation is less than or equal to the adjustment dead zone, the adjustment deviation is set to zero; when the absolute value of the original adjustment deviation is greater than the adjustment dead zone, the adjustment deviation is set to the signed difference between the original adjustment deviation and the adjustment dead zone.
[0031] Set adjustment dead zone Megawatts. Original adjustment deviation. Corrected adjustment deviation for:
[0032] in This is a sign function. Setting a dead zone avoids frequent adjustments caused by measurement noise. This calculation is implemented in the AGC server software using conditional statements.
[0033] S17: Based on the current head, historical head change sequence and preset head-output correction curve of each power station, dynamically correct the fixed output upper limit and fixed output lower limit of each power station, and generate the dynamic output upper limit and dynamic output lower limit of each power station respectively.
[0034] The head-output correction curve is a characteristic curve obtained from turbine model tests, describing the maximum and minimum stable output of the unit under different heads. (Fixed output upper limit) To determine the rated output under the design head, a fixed lower limit for output is established. The minimum output required for stable unit operation is typically 20% of the rated output. Dynamic correction converts the fixed limit to the current head. The corresponding dynamic limits are implemented through the following sub-steps.
[0035] S171: Collect the current head value of each power station and calculate the head change rate and head change acceleration based on the historical head change sequence.
[0036] Let the current water head be The water head a moment ago was Time interval Seconds. Rate of change of water head. The unit is meters per second. Acceleration due to head change. The unit is meters per square second. The historical head change sequence is stored in a circular array of length 10 and updated after each sampling.
[0037] S172: Query the head-output correction curve based on the current head value to obtain the basic correction coefficient.
[0038] The head-output correction curve is stored in a table format in the database. Linear interpolation is used during queries: if the current head... Between and The corresponding correction factors are as follows: and Then the basic correction factor This interpolation can be implemented using function blocks in a PLC, or calculated in a host computer using Python's numpy.interp function.
[0039] S173: The basic correction coefficient is dynamically adjusted according to the rate of change of water head to generate the first correction coefficient.
[0040] The first-order dynamic adjustment formula is ,in This is a first-order gain. When the water head rises rapidly... , Increase the dynamic output limit to fully utilize the incoming water; when the water head decreases , Reduce the dynamic output limit to prevent over-output.
[0041] S174: The first correction coefficient is dynamically adjusted in a second order based on the acceleration of the change in water head to generate the final correction coefficient.
[0042] The second-order dynamic adjustment formula is: ,in It is a second-order gain. Acceleration due to head change. The curvature reflects the change in head. When the acceleration is positive, the correction coefficient is further increased, and when it is negative, the correction coefficient is decreased to compensate for the inertial effect.
[0043] S175: The product of the fixed output upper limit and the final correction coefficient is used as the dynamic output upper limit, and the product of the fixed output lower limit and the final correction coefficient is used as the dynamic output lower limit.
[0044] Dynamic output limit Dynamic output lower limit This dynamic limit changes in real time with the water head and is updated every second.
[0045] S18: Calculate the dynamic positive reserve capacity of each power station based on its current output, dynamic output limit, and unit aging correction factor.
[0046] Dynamic positive reserve capacity reflects the power plant's current capacity to increase output, and its value is affected by the aging of the generating units and the reliability of regulation. This is achieved through the following sub-steps.
[0047] S181: Calculate the difference between the dynamic output limit and the current output, and use this difference as the initial positive reserve capacity.
[0048] Initial positive reserve capacity ,in To contribute to the present.
[0049] S182: Based on the cumulative running time and start-up / shutdown times of each unit in each power station, query the preset aging curve and generate the unit aging correction factor for each power station.
[0050] The aging curve is an empirical curve obtained through statistical analysis of long-term operating data of similar units. The horizontal axis represents the cumulative operating time or number of start-ups and shutdowns, and the vertical axis represents the aging correction factor. The value ranges from 0.7 to 1.0. The longer the cumulative running time and the more start-stop cycles, the smaller the aging correction factor.
[0051] S183: Generate the regulation reliability coefficient for each power station based on its historical regulation success rate.
[0052] Historical adjustment success rate The reliability coefficient refers to the proportion where, over a past period, such as the past 100 regulation commands, the absolute value of the ratio of the actual change in power output to the required change is close to 1. The regulation reliability coefficient ranges from 0.8 to 1.0, with a higher success rate resulting in a coefficient closer to 1.
[0053] S184: Multiply the initial positive reserve capacity by the unit aging correction factor and the regulation reliability coefficient to obtain the dynamic positive reserve capacity.
[0054] Dynamic positive reserve capacity .
[0055] S19: Calculate the dynamic negative reserve capacity of each power station based on its current output, dynamic lower limit of output, and safety margin coefficient.
[0056] Dynamic negative reserve capacity reflects the power station's current capacity to reduce output, and its value is affected by head safety margin and ecological flow constraints. This is achieved through the following sub-steps.
[0057] S191: Calculate the difference between the current output and the lower limit of dynamic output, and use this difference as the initial negative reserve capacity.
[0058] Initial negative reserve capacity .
[0059] S192: Generate head safety margin coefficients based on the difference between the current head and the minimum operating head of each power station.
[0060] Let the minimum operating head be Rated head is Safety margin coefficient of water head The value ranges from 0 to 1. The smaller the margin, the closer the coefficient is to 0, indicating that it is not advisable to reduce the output significantly.
[0061] S193: Based on the downstream ecological flow requirements of each power station, generate ecological flow constraint coefficients.
[0062] Let the minimum ecological discharge flow be The current outflow is The rated discharge flow rate is Ecological flow constraint coefficient The value ranges from 0 to 1.
[0063] S194: Multiply the initial negative reserve capacity by the head safety margin coefficient and the ecological flow constraint coefficient to obtain the dynamic negative reserve capacity.
[0064] Dynamic negative reserve capacity .
[0065] S110: Perform a consistency check on the calculated dynamic positive reserve capacity and dynamic negative reserve capacity. The consistency check includes checking whether the sum of the dynamic positive reserve capacity and the dynamic negative reserve capacity is less than or equal to the preset total regulation capacity threshold. If not, scale the two proportionally so that the sum of the scaled two equals the total regulation capacity threshold.
[0066] Total regulation threshold Typically, a certain percentage, such as 80%, is taken as the sum of the rated capacities of all power plants. Let the sum of the dynamic positive reserve capacities of all power plants be... The sum of dynamic negative reserve capacity is .like Then the scaling ratio The scaled-down dynamic positive reserve capacity of each power station is: Dynamic negative reserve capacity is .
[0067] S2: Based on the sign of the regulation deviation, the dynamic positive reserve capacity and dynamic negative reserve capacity of each power station, and the coupling relationship between the outflow of the upstream power station and the head of the downstream power station, the regulation deviation is distributed in multiple levels to generate the distribution regulation amount of each power station.
[0068] Judging adjustment deviation The positive or negative sign. If >0 indicates that the total output needs to be increased. <0 indicates that the total output needs to be reduced. Then, based on the dynamic positive or negative reserve capacity of each power station and considering the hydraulic coupling relationship between upstream and downstream, the regulation deviation is distributed to each power station using a proportional allocation method. The specific allocation method is shown in S21 to S24.
[0069] S21: Construct a transfer function to determine the impact of upstream power plant outflow on downstream power plant head. The process of generating this transfer function includes obtaining the outflow rate of the upstream power plant, the reservoir capacity characteristic curve of the downstream power plant, and the delay parameters of the water conveyance channel. Based on these parameters, a mapping relationship between the change in upstream outflow rate and the change in downstream head is generated.
[0070] The transfer function is used to describe the change in the input quantity, i.e., the upstream outflow quantity. With output, i.e., changes in downstream head The dynamic relationship between them. A first-order inertial plus pure time-delay model is adopted:
[0071] in This is the steady-state gain, expressed in meters per cubic meter per second. The delay time is in seconds. The time constant is expressed in seconds. The Laplace operator is used. Model parameters were obtained through field step experiments combined with least squares identification, using the prediction error method in the MATLAB System Identification Toolbox. The identified model parameters are stored in the configuration file of the AGC server.
[0072] S22: When the adjustment deviation is positive, the predicted increment of the downstream power station head due to the increase in upstream output is calculated according to the transfer function, and the dynamic positive reserve capacity of the downstream power station is adjusted downward based on the predicted increment, with the adjustment magnitude being proportional to the predicted increment.
[0073] when At that time, let the increased power output allocated to the upstream power station be... Through the head-to-flow conversion coefficient Will Converted to upstream outflow increment The head-to-flow conversion factor is: ,in The turbine efficiency was obtained by interpolation of the efficiency characteristic surface provided by the manufacturer. The current water head. The input transfer function is used to obtain the predicted downstream head increment through inverse Laplace transform. .
[0074] Lowering the amount ,in The adjustment factor is downward. The corrected dynamic positive reserve capacity of downstream power plants is: This correction can prevent over-regulation caused by downstream units passively increasing their output due to an increase in downstream head.
[0075] S23: When the adjustment deviation is negative, the predicted reduction in the head of the downstream power station due to the decrease in upstream output is calculated based on the transfer function, and the dynamic negative reserve capacity of the downstream power station is adjusted upward based on the predicted reduction. The upward adjustment is proportional to the predicted reduction.
[0076] when At that time, let the reduced output power allocated to the upstream power station be... Through the head-to-flow conversion coefficient Convert it into upstream outflow reduction To achieve real-time online computation, the continuous transfer function needs to be discretized. A bilinear transform is employed, with a sampling period of... In seconds, the discrete difference equation is obtained:
[0077] in This is the predicted value of the downstream head change at the current moment. Input for changes in upstream outflow. For pure time delay steps, the coefficient is... Calculated using the bilinear transformation formula. As input Substituting into the difference equation, the predicted value of downstream head change is recursively calculated. And take the absolute value. Increase range. ,in This is an upward adjustment factor. The corrected dynamic negative reserve capacity of downstream power plants is... This correction compensates for the passive reduction in output by downstream units due to a drop in head, resulting in more accurate power allocation. The entire calculation process is performed once per regulation cycle, with a calculation delay controlled within 10 milliseconds.
[0078] S24: Based on the corrected dynamic positive reserve capacity or the corrected dynamic negative reserve capacity, the adjustment deviation is allocated to each power station using a proportional allocation method, generating the allocated adjustment amount for each power station. The proportional allocation method means that the allocated adjustment amount undertaken by each power station is the same as the proportion of its corrected dynamic reserve capacity to the total dynamic reserve capacity.
[0079] for , No. The allocation and regulation amount for each power station is:
[0080] for , No. The allocation and regulation amount for each power station is:
[0081] The allocation results are in megawatts, rounded to one decimal place. This calculation is performed using floating-point arithmetic in the AGC server software.
[0082] S3: Determine the unit regulation sequence based on the comprehensive regulation capability index of each power station's internal units, and generate the unit regulation sequence.
[0083] The comprehensive regulation capacity index is a comprehensive evaluation value used to measure the priority of each unit in responding to output regulation. This step calculates the index for all units within each power station, and then sorts them from high to low to generate a unit regulation sequence. The index comprehensively considers factors such as the unit's remaining regulation capacity, efficiency characteristics, vibration zone constraints, and cumulative operating time. The specific calculation method is shown in S31 to S36.
[0084] S31: Based on the difference between the current output and the upper limit of dynamic output or the difference between the current output and the lower limit of dynamic output of each unit, the remaining regulation capacity value of each unit is calculated.
[0085] For scenarios that require increased output, i.e. Remaining regulation capacity value For scenarios where reduced output is required, i.e. , .
[0086] S32: Based on the unit flow power generation of each unit within the preset output range, the efficiency characteristic coefficient of each unit is calculated.
[0087] Power generation per unit flow The unit is megawatts per cubic meter per second, of which This refers to the generator's power generation flow rate. Efficiency characteristic coefficient. , This represents the average power generation per unit flow of all operating units within the same power station.
[0088] S33: Based on the historical vibration test data of each unit, determine the boundary value of the output vibration range that each unit should avoid, and calculate the reciprocal of the distance between the current output and the nearest vibration range boundary as the vibration zone penalty factor.
[0089] Let the vibration interval be... Current output Distance from the boundary of the nearest vibration interval Vibration zone penalty factor ,in Megawatts are used to prevent small positive numbers from being divided by zero.
[0090] S34: Based on the continuous operating time of each unit before the current adjustment cycle, calculate the cumulative operating time of each unit, and generate a fatigue correction coefficient according to the preset operating time fatigue curve.
[0091] Continuous runtime The unit is hours. Fatigue correction factor. .
[0092] S35: The remaining regulating capacity value, efficiency characteristic coefficient, vibration zone penalty factor, and fatigue correction coefficient are weighted and summed to obtain the comprehensive regulating capacity index of each unit. The weight of the vibration zone penalty factor is set to different values when the regulating deviation is positive and negative.
[0093] Comprehensive regulation capacity index Basic weight .when and At that time, Temporarily increase to 0.5, while proportionally reducing other weights; when and The same treatment applies at the same time.
[0094] S36: Generate the unit control sequence from high to low based on the comprehensive control capability index.
[0095] All units are classified according to the comprehensive regulation capacity index Sort from largest to smallest to generate an ordered list.
[0096] S4: According to the unit regulation sequence, the unit is controlled to perform output regulation in sequence by step approximation method, and the output limit, vibration zone boundary and head change feedback of each unit are monitored in real time during the regulation process until the allocated regulation amount of each power station is completed.
[0097] The step-by-step approximation method is an iterative adjustment strategy, which involves applying a small output adjustment step to each unit each time, and then proceeding to the next adjustment after the system stabilizes. This step controls each unit to increase or decrease its output according to the unit adjustment sequence generated in S3. See S41 to S45 for specific implementation details.
[0098] S41: Select the unit with the highest current comprehensive regulation capacity index from the unit regulation sequence as the current regulating unit.
[0099] S42: Determine the adjustable output range of the current regulating unit based on the current power station's allocated regulation amount, the current regulating unit's dynamic output upper limit and dynamic output lower limit, and the current regulating unit's vibration range boundary value.
[0100] Adjustable output range .like Then skip that unit.
[0101] S43: Controls the current regulating unit to gradually increase or decrease its output within the adjustable output range according to a dynamic step size.
[0102] Adjustable output range length Basic step size Dynamic step size ,in , This represents the current rate of change of head. An increase or decrease in output power is sent based on the sign of the adjustment deviation.
[0103] S44: After each long adjustment step is completed, the actual output of the current regulating unit is re-acquired, and it is determined whether the boundary of the adjustable output range has been reached or whether the allocated adjustment amount undertaken by the current unit has been completed.
[0104] S45: If the boundary has been reached or the allocated regulation amount has been completed, stop the regulation of the current regulating unit, record the actual output of the current unit as the initial output state of the next unit, and then switch to the next unit in the unit regulation sequence. Repeat S42 to S44 until the allocated regulation amount of each power station is completed.
[0105] S5: During the output regulation process, a safe and feasible domain is constructed based on the flow rate change rate and head change rate of the water conveyance channel, and dynamic safety limits are imposed on the regulation rate and regulation increment according to real-time hydraulic parameters.
[0106] This step is executed in parallel during the adjustment process, and safety interlocking is performed with a higher priority than AGC. See S51 to S55 for the specific implementation.
[0107] S51: Real-time acquisition of flow rate change rate and head change rate in the water conveyance channel. Flow rate change rate Rate of change of water head ,Pick Seconds. Among them, This represents the measured flow rate of the water conveyance channel at the current moment. This is the measured flow rate at the previous sampling time. This is the measured head value at the current moment. This is the measured water head value at the previous sampling time. The sampling interval is denoted as .
[0108] S52: Setting the upper limit threshold for flow rate change based on hydraulic design parameters and the upper limit threshold of the rate of change of water head And set the first warning interval. Second warning interval .
[0109] S53: Constructing a secure feasible domain .
[0110] S54: Calculate the predicted flow rate of change and predicting the rate of change of head ,in To adjust the rate, This is an empirical coefficient. The value range is 0.5-2.0. The value ranges from 0.01 to 0.10, and the specific value is related to the hydraulic design parameters of the power station. It can be adjusted based on measured data during on-site commissioning. This represents the change in output power at the current adjustment step size. This is the execution time for the current adjustment step size.
[0111] S55: Implement graded security restrictions. If or If the price falls into the warning zone, reduce the adjustment rate to 50% of the original rate, and limit the adjustment increment to 60% of the original increment. If the price exceeds the safe and feasible range, immediately pause the adjustment and reverse the adjustment. After pausing, reassess every 20 seconds, and resume adjustment at the reduced rate and step size once the price falls back into the warning zone.
[0112] S6: Based on the predicted load change trend and inflow flow change trend, feedforward compensation correction is performed for the regulation cycle that has not yet occurred, and the reserve capacity allocation ratio of each power station is adjusted in advance.
[0113] This step obtains the future load curve from the power grid dispatching system and the inflow forecast from the upstream hydrological forecasting system, and uses a rolling time-domain optimization algorithm to calculate the feedforward compensation. See S61 to S64 for details.
[0114] S61: The sliding window averaging method is used to extract load change trend characteristics. Four load points within the next 15 minutes are selected. Calculate the slope of the linear regression. , as the rate of load change; in, For the time corresponding to the j-th load point, Let j be the total active power load value corresponding to the j-th load point. The arithmetic mean of the times at the four selected load points. This is the arithmetic mean of the load values at the four selected load points.
[0115] S62: Predicted value of the trend of production water flow using exponential smoothing method. ,Pick The predicted increase in water inflow over the next 2 hours is... . This represents the exponentially smoothed value of the inflow rate at the current time t. This is the smoothed value of the inflow flow index from the previous moment. For smoothing coefficients, This is the predicted increase in water flow over the next 2 hours. This is a smoothed value for the predicted inflow rate of the second hour in the future, obtained by extrapolation using the exponential smoothing method.
[0116] S63: The feedforward compensation amount is calculated using a rolling time-domain optimization algorithm. The prediction time domain is set. Each cycle controls the time domain. Objective function The constraints are: .
[0117] in, The total predicted output of the cascade hydropower stations in the k-th prediction period is given. This refers to the total active power command predicted by the scheduling system during the k-th prediction period. This represents the reserve capacity adjustment for the i-th power plant in the k-th forecast period. A positive value indicates an increase in output, and a negative value indicates a decrease in output. Let be the dynamic positive reserve capacity of the i-th power station. Let be the dynamic negative reserve capacity of the i-th power station.
[0118] Solving the quadratic programming problem yields the optimal reserve capacity adjustment sequence, and the first value is taken as the feedforward compensation amount. It is updated on a rolling basis after each cycle ends.
[0119] S64: Add the feedforward compensation to the current adjustment deviation allocation to generate the feedforward corrected allocation adjustment. , Let be the feedforward adjustment coefficient, taken as... , This represents the final allocation adjustment for the i-th power station after feedforward correction.
[0120] S7: Based on the deviation between the actual total output and the dispatch command after the end of each regulation cycle, as well as the deviation between the actual hydraulic parameters and the predicted hydraulic parameters, multivariate corrections are made to the regulation deviation allocation method, the unit regulation sequence, and the boundary of the safe and feasible domain.
[0121] This step calculates two deviations after each adjustment period and inputs them into the multivariate adaptive correction model. See S71 to S74 for details.
[0122] S71: Calculate the periodic output deviation .
[0123] S72: Calculate hydraulic prediction deviation .
[0124] S73: Will and Input a multivariate adaptive correction model. Set the first correction coefficient. Second correction coefficient vector Third correction factor The initial values are all 1.0. Define the loss function. The gradient descent method is used to update the correction coefficients synchronously. The learning rate The gradient is calculated using numerical differencing. When three consecutive periods... Megawatt-hour freeze update.
[0125] S74: Based on the updated correction coefficients, adjust the allocation method of adjustment deviation, the unit adjustment sequence, and the boundary of the safe and feasible domain for the next cycle.
[0126] S91: Mark the remaining regulation capacity that the power station can no longer handle as load to be transferred.
[0127] S92: Calculate the impact factor of each unit increase in power output of the upstream power station on the head of the downstream power station. ,in For the steady-state gain of the transfer function, This is the head-to-flow conversion coefficient.
[0128] S93: When At that time, the downstream water head rose. Estimated output loss ,in This is the head-output sensitivity coefficient. Additional compensation adjustment amount. The actual increase in upstream power output was .
[0129] S94: When At that time, the downstream water head dropped. Estimated reduction in output Additional relief adjustment amount The actual reduction in upstream power output was... .
[0130] S95: Re-verify the dynamic reserve capacity of upstream power plants. If it cannot simultaneously handle the original allocation adjustment and the load to be transferred, the excess portion will be re-allocated to the more upstream power plants.
[0131] The above steps together constitute a complete implementation plan for the coordinated control method of cascade hydropower stations. Through these steps, it is possible to achieve reasonable power output allocation and rapid, stable regulation while meeting hydraulic engineering safety constraints and generator unit safety constraints.
[0132] like Figure 2 As shown in some embodiments, a coordinated control system for a cascade hydropower station includes: The data acquisition module 201 is used to collect real-time operating status parameters, hydraulic environment parameters, and total active power commands issued by the dispatching system for each power station in the cascade hydropower station. The dynamic reserve calculation module 202 is used to correct the upper and lower limits of output based on the current head, and to calculate the dynamic positive reserve capacity and dynamic negative reserve capacity based on the corrected upper and lower limits of output and the current output. The hydraulic coupling distribution module 203 is used to distribute the adjustment deviation in multiple levels according to the sign of the adjustment deviation, the dynamic reserve capacity of each power station, and the transfer function of the influence of the outflow of the upstream power station on the head of the downstream power station, and to generate the distribution adjustment amount of each power station. The unit sequencing module 204 is used to calculate the comprehensive regulation capacity index based on the unit's remaining regulation capacity, efficiency characteristic coefficient, vibration zone penalty factor and fatigue correction coefficient, and generate the unit regulation sequence according to the comprehensive regulation capacity index. The adjustment execution module 205 is used to control the unit to perform output adjustment within the adjustable output range according to the unit adjustment sequence and using a step-approximation method. The adjustable output range is dynamically determined based on the dynamic output upper and lower limits and the vibration range boundary. The safety constraint module 206 is used to construct a safe and feasible region based on the flow rate change rate and head change rate of the water conveyance channel, and to implement graded safety constraint measures when the regulation process exceeds the safe and feasible region, including reducing the regulation rate, limiting the regulation increment, suspending regulation, and reverse back-off regulation; The feedforward prediction module 207 is used to generate feedforward compensation based on the load change trend and the inflow flow change trend, using a rolling time-domain optimization algorithm, and to adjust the reserve capacity allocation ratio of each power station in advance. The correction module 208 is used to perform multivariate adaptive correction of the allocation correction coefficient, sorting weighting coefficient and safe feasible domain boundary scaling coefficient based on the periodic output deviation and hydraulic prediction deviation, using the gradient descent method.
[0133] It is understandable that the modules recorded in the coordinated control system of this cascade hydropower station are consistent with the reference... Figure 1 The steps described in the coordinated control method for cascade hydropower stations correspond to those steps in this paper. Therefore, the operations, characteristics, and beneficial effects described above for the coordinated control method for cascade hydropower stations also apply to the coordinated control system for cascade hydropower stations and its constituent modules, and will not be repeated here.
[0134] The above description is merely an explanation of some preferred embodiments of the present invention and the technical principles employed. Those skilled in the art should understand that the scope of the invention as described in the embodiments of the present invention is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the above-described inventive concept. For example, technical solutions formed by substituting the above-described features with (but not limited to) technical features with similar functions disclosed in the embodiments of the present invention.
Claims
1. A method for coordinated control of cascade hydropower stations, characterized in that, Includes the following steps: S1: Collect real-time operating status parameters, hydraulic environment parameters, and total active power commands issued by the dispatching system for each power station in the cascade hydropower station. Calculate the current total output, regulation deviation, dynamic positive reserve capacity, and dynamic negative reserve capacity based on the collected data. S2: Construct a transfer function to determine the impact of upstream power plant outflow on downstream power plant head. The generation process of the transfer function includes: obtaining the outflow rate of the upstream power plant, the reservoir capacity characteristic curve of the downstream power plant, and the delay parameters of the water conveyance channel; and fitting the above parameters to generate a mapping relationship between the change in upstream outflow rate and the change in downstream head. When the adjustment deviation is positive, calculate the predicted increment of the downstream power plant head increase due to the increase in upstream output based on the transfer function, and adjust the dynamic positive reserve capacity of the downstream power plant downward based on the predicted increment, with the downward adjustment magnitude being proportional to the predicted increment. When the adjustment deviation is negative, the predicted reduction in head of the downstream power station due to the decrease in upstream output is calculated according to the transfer function. Based on the predicted reduction, the dynamic negative reserve capacity of the downstream power station is adjusted upward, with the adjustment amount proportional to the predicted reduction. According to the corrected dynamic positive reserve capacity or the corrected dynamic negative reserve capacity, the adjustment deviation is allocated to each power station using a proportional allocation method to generate the allocated adjustment amount for each power station. The proportional allocation method is as follows: the allocated adjustment amount undertaken by each power station is the same as the proportion of its corrected dynamic reserve capacity to the total dynamic reserve capacity. When the distribution adjustment of a power station reaches its dynamic output upper or lower limit, and the predicted change in the impact of the outflow from the upstream power station on the power station's head exceeds a preset impact threshold, a residual load reallocation step is executed, including: The remaining regulation capacity that the power station can no longer handle is marked as load to be transferred; the impact factor of each unit increase in output of the upstream power station on the head of the downstream power station, and the impact factor of each unit decrease in output of the upstream power station on the head of the downstream power station are calculated according to the transfer function; when the regulation deviation is positive, the load to be transferred is allocated to the upstream power station, and the allocation regulation amount of the upstream power station is adjusted simultaneously, so that the increased output of the upstream power station can partially compensate for the potential output loss of the downstream power station due to the rise in head; when the regulation deviation is negative, the load to be transferred is allocated to the upstream power station, and the allocation regulation amount of the upstream power station is adjusted simultaneously, so that the decreased output of the upstream power station can alleviate the risk of over-output of the downstream power station due to the drop in head; after the remaining load is redistributed, the dynamic reserve capacity of the upstream power station is re-verified to see if it can simultaneously handle its original allocation regulation capacity and the load to be transferred; if not, the excess portion is further recursively allocated to the upstream power station. S3: Determine the unit regulation sequence based on the comprehensive regulation capacity index of each power station's internal units, and generate the unit regulation sequence; S4: According to the unit adjustment sequence, the unit is controlled to perform output adjustment in sequence by step approximation method, and the output limit, vibration zone boundary and head change feedback of each unit are monitored in real time during the adjustment process until the distribution adjustment of each power station is completed. S5: During the output regulation process, a safe and feasible domain is constructed based on the flow rate change rate and head change rate of the water conveyance channel, and dynamic safety limits are imposed on the regulation rate and regulation increment according to real-time hydraulic parameters. S6: Based on the predicted load change trend and inflow flow change trend, feedforward compensation correction is performed for the regulation cycle that has not yet occurred, and the reserve capacity allocation ratio of each power station is adjusted in advance. S7: Based on the deviation between the actual total output and the dispatch command after the end of each adjustment cycle, as well as the deviation between the actual hydraulic parameters and the predicted hydraulic parameters, multivariate corrections are made to the adjustment deviation allocation method, the unit adjustment sequence, and the boundary of the safe and feasible domain.
2. The method of claim 1, wherein, S1 includes: S11: Collect real-time operating status parameters of each power station in the cascade hydropower station. The real-time operating status parameters include the current active power output, current head, current water conveyance channel flow, operating status identifier of each unit, and historical output sequence of each unit. S12: Collect the total active power command issued by the scheduling system, and simultaneously collect the timestamp and valid identifier of the total active power command. S13: Perform data preprocessing on the current active power output of each power station. The data preprocessing includes: removing outliers that are significantly beyond the equipment range, interpolating and filling missing data, and performing first-order low-pass filtering on the data after removal and filling to generate the filtered active power output of each power station. S14: The filtered active power output of each power station is summed and the sum is used as the current total output. S15: Perform validity verification on the total active power instruction. The validity verification includes: checking whether the instruction validity identifier is in a valid state, checking whether the difference between the timestamp and the current system time is less than the preset maximum allowable delay. If the verification passes, the total active power instruction is used as a valid instruction for subsequent calculations. If the verification fails, the valid instruction of the previous cycle is used and an invalid instruction alarm is generated. S16: Calculate the difference between the effective command and the current total output, use the difference as the original adjustment deviation, and perform dead-zone correction on the original adjustment deviation based on a preset adjustment dead zone to generate an adjustment deviation for subsequent allocation; the specific method of the dead-zone correction is as follows: when the absolute value of the original adjustment deviation is less than or equal to the adjustment dead zone, the adjustment deviation is set to zero; when the absolute value of the original adjustment deviation is greater than the adjustment dead zone, the adjustment deviation is set as the signed difference between the original adjustment deviation and the adjustment dead zone. S17: Based on the current head, historical head change sequence and preset head-output correction curve of each power station, dynamically correct the fixed output upper limit and fixed output lower limit of each power station, and generate the dynamic output upper limit and dynamic output lower limit of each power station respectively. S18: Calculate the dynamic positive reserve capacity of each power station based on its current output, dynamic output limit, and unit aging correction factor. S19: Calculate the dynamic negative reserve capacity of each power station based on its current output, dynamic output lower limit, and safety margin coefficient. S110: Perform a consistency check on the calculated dynamic positive reserve capacity and the dynamic negative reserve capacity. The consistency check includes: checking whether the sum of the dynamic positive reserve capacity and the dynamic negative reserve capacity is less than or equal to a preset total regulation capacity threshold. If not, scale the two proportionally so that the sum of the scaled two equals the total regulation capacity threshold.
3. The method of claim 1, wherein, S4 includes: S41: Select the unit with the highest current comprehensive regulation capability index from the unit regulation sequence as the current regulating unit; S42: Determine the adjustable output range of the current regulating unit based on the current power plant's allocated regulation amount, the current regulating unit's dynamic output upper limit and dynamic output lower limit, and the current regulating unit's vibration range boundary value; S43: Control the current regulating unit to gradually increase or decrease its output within the adjustable output range according to a dynamic step size; S44: After each long adjustment step is completed, the actual output of the current regulating unit is re-acquired, and it is determined whether the boundary of the adjustable output range has been reached or whether the allocated adjustment amount undertaken by the current unit has been completed. S45: If the boundary has been reached or the allocated adjustment amount has been completed, stop the adjustment of the current regulating unit, record the actual output of the current unit as the initial output state of the next unit, and then switch to the next unit in the unit adjustment sequence. Repeat S42 to S44 until the allocated adjustment amount of each power station is completed.
4. The method of claim 1, wherein, S5 includes: S51: Real-time acquisition of the flow rate change rate and head change rate of the water conveyance channel, wherein the flow rate change rate is calculated based on the flow rate change per unit time, and the head change rate is calculated based on the head change per unit time. S52: Based on the hydraulic design parameters, set the upper limit threshold for flow rate change rate and the upper limit threshold for head change rate, and further set the first warning interval and the second warning interval; wherein, the first warning interval is the range of 70% to 90% of the upper limit threshold for flow rate change rate, and the second warning interval is the range of 70% to 90% of the upper limit threshold for head change rate. S53: Construct the safe and feasible region using the upper limit threshold of the flow rate change rate and the upper limit threshold of the head change rate as absolute boundaries; the safe and feasible region is defined as the set of adjustment parameters that simultaneously satisfy the condition that the flow rate change rate is less than or equal to the upper limit threshold of the flow rate change rate and the head change rate is less than or equal to the upper limit threshold of the head change rate. S54: During the power output regulation process, calculate the predicted flow rate change rate and predicted head change rate corresponding to the current regulation rate and the current regulation increment; S55: Implement graded safety restriction measures based on the numerical range of the predicted flow rate change rate and the predicted head change rate.
5. The method of claim 1, wherein, S6 includes: S61: Based on the predicted load curves for multiple future time points issued by the power grid dispatching system, the load change trend feature value is extracted using the sliding window averaging method. The load change trend feature value includes the load increase rate and load decrease rate within the first preset period in the future. S62: Based on the inflow forecast provided by the upstream hydrological forecasting system, the inflow flow change trend prediction value is generated by exponential smoothing method. The inflow flow change trend prediction value includes the inflow increment prediction and inflow reduction prediction within the second preset period. S63: Based on the characteristic value of the load change trend and the predicted value of the local water flow change trend, the feedforward compensation amount for each future regulation cycle is calculated using a rolling time-domain optimization algorithm; S64: The feedforward compensation amount is superimposed on the current adjustment deviation allocation to generate the feedforward corrected allocation adjustment amount.
6. The method of claim 2, wherein, S7 includes: S71: After each adjustment cycle, calculate the cycle output deviation between the actual total output and the scheduling command; S72: Simultaneously calculate the hydraulic prediction deviation between the actual head change and the head change predicted based on the transfer function; S73: Input the periodic output deviation and the hydraulic prediction deviation into the multivariate adaptive correction model, and synchronously iterate and update multiple correction coefficients; S74: Based on the updated correction coefficients, adjust the allocation method of adjustment deviation, the unit adjustment sequence, and the boundary of the safe and feasible domain for the next cycle.
7. The method of claim 1, wherein, S3 includes: S31: Based on the difference between the current output and the upper limit of dynamic output or the difference between the current output and the lower limit of dynamic output of each unit, calculate the remaining regulation capacity value of each unit. S32: Based on the unit flow power generation of each unit within the preset output range, the efficiency characteristic coefficient of each unit is calculated. S33: Based on the historical vibration test data of each unit, determine the boundary value of the output vibration range that each unit should avoid, and calculate the reciprocal of the distance between the current output and the nearest vibration range boundary as the vibration zone penalty factor. S34: Based on the continuous running time of each unit before the current adjustment cycle, calculate the cumulative running time of each unit, and generate a fatigue correction coefficient according to the preset running time fatigue curve. S35: The remaining regulation capacity value, the efficiency characteristic coefficient, the vibration zone penalty factor and the fatigue correction coefficient are weighted and summed to obtain the comprehensive regulation capacity index of each unit. S36: Generate the unit regulation sequence from high to low based on the comprehensive regulation capability index.
8. A coordinated control system for a cascade hydropower plant, implementing the method according to any one of claims 1 to 7, characterized in that, include: The data acquisition module is used to collect real-time operating status parameters, hydraulic environment parameters, and total active power commands issued by the dispatching system for each power station in the cascade hydropower station. The dynamic reserve calculation module is used to correct the upper and lower limits of output based on the current head, and to calculate the dynamic positive reserve capacity and dynamic negative reserve capacity based on the corrected upper and lower limits of output and the current output. The hydraulic coupling distribution module is used to distribute the adjustment deviation in multiple levels according to the sign of the adjustment deviation, the dynamic reserve capacity of each power station, and the transfer function of the influence of the outflow of the upstream power station on the head of the downstream power station, and to generate the distribution adjustment amount of each power station. The unit sequencing module is used to calculate the comprehensive regulation capacity index based on the unit's remaining regulation capacity, efficiency characteristic coefficient, vibration zone penalty factor, and fatigue correction coefficient, and to generate the unit regulation sequence according to the comprehensive regulation capacity index. The adjustment execution module is used to control the unit to perform output adjustment within the adjustable output range according to the unit adjustment sequence and using a step-approximation method. The adjustable output range is dynamically determined based on the dynamic output upper and lower limits and the vibration range boundary. The safety constraint module is used to construct a safe and feasible region based on the rate of change of flow and the rate of change of head in the water conveyance channel, and to implement graded safety constraint measures when the regulation process exceeds the safe and feasible region, including reducing the regulation rate, limiting the regulation increment, suspending regulation, and reverse back-off regulation; The feedforward prediction module is used to generate feedforward compensation based on load change trends and inflow flow trends using a rolling time-domain optimization algorithm, and to adjust the reserve capacity allocation ratio of each power station in advance. The correction module is used to perform multivariate adaptive correction of the allocation correction coefficient, sorting weighting coefficient, and safe feasible domain boundary scaling coefficient based on the periodic output deviation and hydraulic prediction deviation, using the gradient descent method.
Citation Information
Patent Citations
Real-time load intelligent regulation and control method for runoff type cascade hydropower station group
CN120706812A
Cascade hydropower real-time optimization scheduling method capable of quickly responding to net load deviation
CN121124045A
