Multi-working-condition self-adaptive control method under optimized operation of hydroelectric generating set

Abstract

The invention relates to the technical field of hydroelectric generation, and discloses a multi-working-condition self-adaptive control method under optimized operation of a hydroelectric generating set, which comprises the following steps of: dividing working conditions of a hydroelectric generating set system, and taking a hydroelectric generating set speed regulator control performance comprehensive evaluation index and a set damping characteristic evaluation index as two optimization targets; a multi-target particle swarm optimization algorithm is adopted to solve the Pareto front of the system under each working condition, so that the Pareto front under all working conditions is obtained, and a thin-plate spline method is utilized to fit the optimal control parameter at the position with the minimum Euclidean distance from the original point on the front, so that the optimal control parameter curved surface under all working conditions is obtained; according to the method, the safety problem caused by primary frequency modulation of the unit in a traditional hydroelectric generating set control method is solved, the negative damping characteristic is prevented from being formed, meanwhile, the adjusting speed of the hydroelectric generating set is increased, and the system can be more rapidly kept stable and has good control performance.

Description

Technical Field

[0001] The present invention relates to the technical field of hydropower generation, and particularly relates to a multi-condition adaptive control method for optimized operation of a hydropower unit. Background Art

[0002] Hydropower is a clean energy source with a flexible regulation function. When hydropower units deeply participate in grid regulation, they need to operate within a wider range of working conditions. At the same time, with the expansion of the scale of hydropower, higher requirements are put forward for the calculation of optimized dispatching. The requirements of the State Grid for primary frequency regulation are mainly reflected in three indicators: rise time, delay, and regulation time. These indicators ensure that grid-connected power sources can quickly and accurately respond to changes in grid frequency to maintain the frequency stability of the grid, which is crucial for ensuring the safe and stable operation of the grid.

[0003] Currently, with the further increase in the proportion of wind and solar renewable energy in the grid, the traditional control strategies of hydropower units can no longer meet the requirements for dealing with various complex working conditions. In addition, the units need to bear the dual pressures from the fluctuations in the output of clean energy and the random changes in load. The number of operations under non-optimal working conditions has increased significantly, and the frequency of working condition conversion has correspondingly increased, further deteriorating the operating conditions. In addition, the primary frequency regulation performance and damping characteristics of the units are very important factors in the safe operation of the grid. Most studies focus too much on the quickness of regulation when improving the primary frequency regulation performance of the units, which will have some adverse effects on the units. An overly fast regulation speed will cause the servomotor to twitch, increase the fatigue and wear of mechanical components, reduce the reliability of the guide vane movement, and for hydropower stations with short penstocks or without surge chambers, too fast an opening or closing speed of the guide vanes is likely to cause water hammer phenomena, resulting in too large a pressure difference inside and outside the pipe wall, causing deformation and seriously threatening the safe operation of the water conveyance system. In a power system dominated by hydropower, the occurrence of low-frequency oscillations is directly related to the damping characteristics of hydropower units. A large number of studies have shown that the negative damping effect of hydropower units is the main cause of low-frequency oscillations. Therefore, during the grid regulation process, hydropower units should not only ensure a relatively fast regulation speed but also avoid forming negative damping characteristics. Summary of the Invention

[0004] In view of the above deficiencies in the prior art, the present invention provides a multi-condition adaptive control method for optimized operation of a hydropower unit, which is used to solve the safety problems caused by the primary frequency regulation of traditional hydropower units, and at the same time improve the regulation speed of hydropower units and avoid forming negative damping characteristics.

[0005] In order to achieve the above invention purpose, the technical solution adopted by the present invention is as follows:

[0006] A multi-condition adaptive control method for optimized operation of a hydropower unit, comprising the following steps:

[0007] S1. Obtain several operating conditions allowing the hydroelectric unit system to operate, get the turbine transfer coefficient under a single operating condition, and calculate the state - space equation for the operation of the hydroelectric unit system to determine the stability domain of the control parameters;

[0008] S2. Take the stability domain of the control parameters as the search space of the control parameters. Adopt the multi - objective particle swarm optimization method to calculate the particle fitness by introducing the comprehensive evaluation index of the hydroelectric governor control performance and the evaluation index of the unit damping characteristic, so as to optimize the control parameters when the hydroelectric unit system operates under each operating condition and obtain the Pareto front when the hydroelectric unit system operates under each operating condition;

[0009] S3. Judge whether the control parameter optimization has been completed for all operating conditions. If so, obtain the Pareto fronts under all operating conditions and execute step S6; otherwise, execute step S4;

[0010] S4. Obtain the opening and water head when the hydroelectric unit system operates under the current operating condition, get the turbine transfer coefficient under the current operating condition, and execute steps S1 - S2 to obtain and normalize the Pareto front when the hydroelectric unit system operates under the current operating condition;

[0011] S5. Calculate the Euclidean distance between each point on the normalized Pareto front and the origin, take the optimal solution with the minimum Euclidean distance as the optimal control parameter for the current operating condition, and at the same time increment the operating condition counter by 1, and then execute step S3;

[0012] S6. Use the thin - plate spline method to fit the Pareto fronts under all operating conditions to obtain the fitting surface of the optimal control parameters. By real - time detecting the current opening and water head when the hydroelectric unit system operates in the normal working range and based on the fitting surface, adjust and output the optimal value of the current control parameter.

[0013] The present invention has the following beneficial effects:

[0014] A multi - condition adaptive control method for the optimal operation of a hydroelectric unit proposed by the present invention divides the operating conditions of the hydroelectric unit system. Taking the comprehensive evaluation index of the hydroelectric governor control performance and the evaluation index of the unit damping characteristic as two optimization objectives, it uses the multi - objective particle swarm optimization algorithm to solve the Pareto front of the system under each operating condition, thereby obtaining the Pareto fronts under all operating conditions, and uses the thin - plate spline method to fit the optimal control parameters at the place with the minimum Euclidean distance from the origin on the front to obtain the optimal control parameter surface for the full operating conditions, thus solving the safety problem caused by the primary frequency regulation of the unit in the traditional hydroelectric unit control method, while improving the regulation speed of the hydroelectric unit, avoiding the formation of negative damping characteristics, and enabling the system to maintain stability faster and have good control performance. Description of the Drawings

[0015] Figure 1Schematic flow chart of a multi - condition adaptive control method under optimal operation of a hydropower unit proposed by the present invention;

[0016] Figure 2 Schematic diagram of the general damping characteristic curve of a hydropower unit within the low - frequency oscillation range in the embodiment;

[0017] Figure 3 Schematic diagram of the Pareto front and its corresponding optimal control parameters under PCM in the embodiment;

[0018] Figure 4 Schematic diagram of the system time - domain response and damping corresponding to different control parameters under PCM in the embodiment;

[0019] Figure 5 For the proportional regulation coefficient K P Fitting surface schematic diagram of the regulation;

[0020] Figure 6 For the integral regulation coefficient K I Fitting surface schematic diagram of the regulation;

[0021] Figure 7 Schematic diagram of a typical working condition in the embodiment;

[0022] Figure 8 Schematic diagram of the state change of the LM working condition in the water turbine power mode in the embodiment;

[0023] Figure 9 Schematic diagram of the state change of the HM working condition in the water turbine power mode in the embodiment;

[0024] Figure 10 Schematic diagram of the state change of the ML working condition in the water turbine power mode in the embodiment;

[0025] Figure 11 Schematic diagram of the state change of the MH working condition in the water turbine power mode in the embodiment. Detailed implementation manners

[0026] The following describes the detailed implementation manners of the present invention to facilitate those skilled in the art of the present technology to understand the present invention. However, it should be clear that the present invention is not limited to the scope of the detailed implementation manners. For those of ordinary skill in the art of the present technology, as long as various changes are within the spirit and scope of the present invention defined and determined by the appended claims, these changes are obvious, and all inventions and creations using the concept of the present invention are within the scope of protection.

[0027] As Figure 1 shown, a multi - condition adaptive control method under optimal operation of a hydropower unit includes the following steps S1 - S6:

[0028] S1. Obtain several operating conditions that allow the hydro-generator unit system to operate, get the turbine transfer coefficient under a single operating condition, and calculate the state-space equation of the hydro-generator unit system operation to determine the stable region of the control parameters.

[0029] In this embodiment, based on the important condition for system stability that the roots of the characteristic equation of the stable system all have negative real parts, the stable region can be obtained, thereby determining the search space of the control parameters. In addition, the opening range of the target unit under normal operating conditions is 50% - 100%, and the head range is 155m - 225m. Therefore, a new operating condition point is determined every 10% opening and 10m head, and a total of 48 operating condition points can be obtained. At this time, the operating condition counter is set to 0, and the division of the operating conditions of the hydro-generator unit system is completed.

[0030] Specifically, in step S1, the control parameters are the proportional regulation coefficient K P , the integral regulation coefficient K I .

[0031] Specifically, the state-space equation of the hydro-generator unit system operation in step S1 is:

[0032]

[0033] Among them, Δdx′ represents the intermediate variable related to differentiation, represents the first derivative of the intermediate variable related to differentiation, Δix represents the integral variable, represents the first derivative of the integral variable, Δy represents the change in the servomotor stroke, represents the first derivative of the change in the servomotor stroke, Δh′ represents the intermediate variable related to the head, represents the first derivative of the intermediate variable related to the head, Δω represents the change in the angular velocity of the hydro-generator unit, represents the first derivative of the change in the angular velocity of the hydro-generator unit, Δω r represents the change in the given rotational speed, ω0 represents the initial angular velocity, E″ q represents the q-axis subtransient electromotive force, E″ d represents the d-axis subtransient electromotive force, ΔE″ q represents the change in the q-axis subtransient electromotive force, ΔE″ d represents the change in the d-axis subtransient electromotive force, represents the first derivative of the change in the d-axis subtransient electromotive force, ΔE q ′ represents the change in the q-axis transient electromotive force, represents the first derivative of the change in the q-axis transient electromotive force, ΔE f represents the change in the excitation electromotive force, represents the first derivative of the change in the excitation electromotive force, K D represents the differential regulation coefficient, K Irepresents the integral regulation coefficient, K P ′ represents the intermediate variable of the proportional regulation coefficient, K a represents the equivalent excitation system gain, δ represents the generator power angle, Δδ represents the change in the generator power angle, represents the first derivative of the change in the generator power angle, X d represents the d-axis synchronous reactance, X′ d represents the d-axis transient reactance, X″ d represents the d-axis subtransient reactance, X q represents the q-axis synchronous reactance, X″ q represents the q-axis subtransient reactance, ΔP er represents the change in the given active power, T j represents the inertia time constant of the hydro-generator unit, T d ′0 represents the d-axis open-circuit transient time constant, T″ d0 represents the d-axis open-circuit subtransient time constant, T″ q0 represents the q-axis open-circuit subtransient time constant, T r represents the equivalent time, T 1v represents the time constant of the differential link, T y represents the response time constant of the main servomotor, T w represents the water inertia time constant, e p represents the regulation rate of the hydro-generator unit, e h represents the water turbine torque-head transfer coefficient, e qy represents the water turbine torque-guide vane opening transfer coefficient, e qh represents the water turbine flow-head transfer coefficient, e qx represents the water turbine flow-speed transfer coefficient, ΔV gr represents the change in the given terminal voltage, K 1,B 、K 2,B 、K 3,B 、K 4,B 、K 5,B 、K 6,B 、K 7,B 、K 8,B 、K 9,B 、K 10,B 、K 11,B 、K 12,B 、K 13,B respectively represent different intermediate variables.

[0034] In this embodiment,

[0035] K 12,B =-1, where X d∑ =Xl +X″ d ,X q∑ =X l +X″ q ,where sin represents the sine function, cos represents the cosine function, V s represents the infinite bus voltage, V d represents the d-axis component of the terminal voltage, V q represents the q-axis component of the terminal voltage, V g represents the terminal voltage, I q represents the q-axis component of the stator current, I d represents the d-axis component of the stator current, E y represents the transfer function of the turbine torque to the guide vane opening, e x represents the transfer coefficient of the turbine torque to the speed, X l represents the sum of the reactances of the transformer and the line.

[0036] S2. Take the stability domain of the control parameters as the search space of the control parameters. Adopt the multi-objective particle swarm optimization method to calculate the particle fitness by introducing the comprehensive evaluation index of the control performance of the hydro-generator governor and the evaluation index of the unit damping characteristic, so as to optimize the control parameters of the hydro-generator unit system under each working condition and obtain the Pareto front of the hydro-generator unit system under each working condition.

[0037] Specifically, the Pareto front in step S2 is the set of optimal control parameters under multi-objective optimization.

[0038] Specifically, the comprehensive evaluation index of the control performance of the hydro-generator governor in step S2 is:

[0039]

[0040] where, H OF represents the comprehensive evaluation index of the control performance of the hydro-generator governor, represents the absolute error time integral index, e(t) represents the error value between the expected output and the actual output of the hydro-generator unit system at the t-th moment, F P1 represents the first penalty function, F P2 represents the second penalty function, if and else both represent judgment conditions, N B represents the penalty function value, T S represents the adjustment time, that is, the shortest time required for the response curve to reach and no longer exceed the error band, T Smax represents the maximum allowable adjustment time, T R represents the rise time, that is, the time required for the response to rise to the final value for the first time, T Rmin represents the shortest allowable rise time, A R represents the overshoot, A ODenote the overshoot as A Rmax Denote the maximum allowable undershoot as A Omax Denote the maximum allowable overshoot as A S Denote the disturbance quantity, which is the absolute value of the difference between the steady-state value before disturbance and the steady-state value after disturbance. Let C(t) denote the response curve, and min and max denote taking the minimum value and the maximum value respectively.

[0041] In this embodiment, the comprehensive evaluation index H of the governor control performance of the hydropower unit OF Is a comprehensive evaluation index of control performance considering overshoot and undershoot limits, enabling the unit to meet the requirements of fast primary frequency regulation while ensuring the safety and stability of its own structure; the calculation formula of this index is: Among them, the first penalty function F P1 , Considering that the rise time of the target state variable caused by the change of the guide vane opening in the transient process cannot be too fast to exceed the speed limit of mechanical components, but also cannot be too long to lose the regulation ability, while the second penalty function F P2 Constrains the overshoot in the regulation process and the power undershoot unique to the hydropower unit, and their calculation formulas are respectively: Among them, T S Denotes the regulation time, that is, the shortest time required for the response curve to reach and no longer exceed the error band. Here, the error band takes ±2% of the final value, and T R Denotes the rise time, that is, the time required for the response to rise to the final value for the first time (when there is no overshoot, it is the time required for the response to rise from 10% of the final value to 90% of the final value), A S Denotes the disturbance quantity, which is the absolute value of the difference between the steady-state value before disturbance and the steady-state value after disturbance. Among them, the steady-state value before disturbance is C(0), and the steady-state value after disturbance is C(∞); in addition, the above formulas All regard the initial state value of the system as 0 and the disturbance direction is upward; while in actual measurement, the initial state is often not 0, and the initial state can be adjusted to zero through translation operation; therefore, when the disturbance direction is downward, the maximum value function in formula Needs to be replaced by the minimum value function.

[0042] Specifically, the evaluation index of the unit damping characteristic in step S2 is:

[0043]

[0044] Among them, D OF Denotes the evaluation index of the damping characteristic of the hydropower unit, D s Denotes the overall damping value of the unit, f denotes the oscillation frequency, and D t (f) denotes the damping of the unit corresponding to different oscillation frequencies f, and real[·] denotes the function of obtaining the real part, and s denotes the Laplace operator. represents the transfer function of the output turbine torque with respect to the input rotational speed, where j represents the imaginary unit and ω d represents the oscillating angular velocity.

[0045] In this embodiment, during the power grid regulation process, the hydroelectric generating unit should not only ensure a fast regulation speed but also avoid forming a negative damping characteristic. Taking D OF as an index to optimize the damping characteristic can not only ensure the stability of the unit but also not significantly affect the regulation performance of the system transient process; at different oscillation frequencies, the damping of the hydroelectric generating unit can be calculated by the formula where can be obtained by converting the above state-space equation, that is, using the ss2tf function in MATLAB software for conversion; in the low-frequency oscillation range, the damping characteristic curve of the hydroelectric generating unit generally presents a shape as Figure 2 shown; however, for hydroelectric generating units, the oscillations occurring under poor regulation parameters are mostly ultra-low-frequency oscillations; therefore, when evaluating the damping characteristic of hydroelectric generating units, the ultra-low-frequency band (0 Hz to 0.1 Hz) and a part of the low-frequency band (0.1 Hz to 0.2 Hz, as the margin of the ultra-low-frequency band) of the oscillation are selected for analysis, where from Figure 2 it can be seen that part of the damping of the water resistance is positive and part is negative in the ultra-low-frequency band. Therefore, the frequency of the first intersection point of the damping characteristic curve and the horizontal axis of y = 0 is defined as the starting frequency f0, and correspondingly, the concerned frequency band is B = (f0 to 0.2) Hz. It should be noted that if the damping of the unit is positive at different oscillation frequencies, it means that the damping characteristic curve has no intersection with the horizontal axis of y = 0, and in this case, the starting frequency can be regarded as 0; within this frequency band, the calculation method of the overall damping characteristic of the unit is as shown in the formula D s = ∑ f∈B D t (f); in summary, from the previous analysis, it can be seen that in engineering, it is desired that D s is a small negative damping or positive damping, that is, the absolute value of D s is as small as possible; when the unit shows a large negative damping, the system tends to oscillate; while when the unit shows a large positive damping, the system response speed slows down. Therefore, the damping characteristic evaluation index as shown in the formula is adopted. The damping characteristic optimized by this index can not only ensure the stability of the unit but also not significantly affect the regulation performance of the system transient process.

[0046] Specifically, step S2 specifically includes S21 - S29:

[0047] S21. Take the stable domain of the control parameters as the search space of the control parameters, and set the particle swarm size, the maximum number of iterations, the inertia weight, the learning factor, and the external particle swarm size, and initialize the particle swarm to obtain the original particle swarm.

[0048] S22. Calculate the fitness of each particle in the original particle swarm, specifically as follows:

[0049] Introduce the first objective function and the second objective function, where the first objective function is the comprehensive evaluation index of the control performance of the hydro-generator governor, and the second objective function is the evaluation index of the unit damping characteristic. And use the first objective function value and the second objective function value as the fitness of each particle in the original particle swarm.

[0050] In this embodiment, the fitness includes two objectives, namely the first objective function and the second objective function. Specifically: According to the time-domain response C(t) of the step given signal by the state-space equation simulation system, and then calculate the first objective function value according to the comprehensive evaluation index of the control performance of the hydro-generator governor. At the same time, convert the state-space equation into a transfer function, and calculate the second objective function value according to the evaluation index of the unit damping characteristic.

[0051] S23. Take the initial position of the particles in the original particle swarm as the historical optimal position of the particles.

[0052] S24. Insert the first non-dominated particles in the original particle swarm into the external particle swarm, determine the grid where the first non-dominated particles in the external particle swarm are located, and count the total number of the first non-dominated particles in each grid, calculate the selection probability of each grid, and at the same time set the current iteration number to 1.

[0053] Among them, the first non-dominated particle is a particle in the original particle swarm whose first objective function value and second objective function value are both less than or equal to the corresponding first objective function value and second objective function value of other particles, and is at least smaller in the first objective function value or the second objective function value.

[0054] S25. Judge whether the current iteration number reaches the maximum iteration number. If so, end the search and output the optimal solution. Otherwise, execute step S26; where the optimal solution is the Pareto front.

[0055] S26. Based on the selection probability of each grid, use the roulette wheel method to select a grid, and randomly select a first non-dominated particle from the external particle swarm in this grid. Take the position of this first non-dominated particle as the global optimal position of the particle swarm, update the positions and velocities of all particles in the original particle swarm, and correct the velocities and positions of the particles according to the boundary constraints.

[0056] In this embodiment, the principle of the selected roulette wheel method is: If a certain part of the disk occupies a large area, that is, the probability is large, then the chance of being selected is greater.

[0057] Specifically, the formulas for updating the positions and velocities of all particles in the original particle swarm in step S26 are as follows:

[0058]

[0059] Among them, v i (t+1) represents the velocity of particle i at time t+1, v i (t) represents the velocity of particle i at time t, w represents the inertia weight, c1 and c2 represent acceleration constants, r1 and r2 represent random numbers between 0 and 1, and pbest i represents the historical optimal position of particle i, gbest represents the global optimal position of the particle swarm, and x i (t+1) represents the position of particle i at time t+1, x i (t) represents the position of particle i at time t.

[0060] In this embodiment, the inertia weight w is used to control the influence of the previous speed, and the acceleration constants c1 and c2 are used to adjust the attraction of the individual and global optimality respectively; in addition, when the particle speed and position are updated using the above formula, the calculated position and speed of each particle also need to be bounded, and the specific process is as follows:

[0061] Specifically, the specific process of correcting the velocity and position of the particle according to the boundary constraint in step S26 is as follows:

[0062] First, according to the set upper and lower limits of particle speed, determine whether the current speed of the particle is greater than the set maximum speed. If so, correct the current speed of the particle to the set maximum speed. Otherwise, determine whether the current speed of the particle is less than the set minimum speed. If so, correct the current speed of the particle to the set minimum speed. Otherwise, keep the current speed of the particle unchanged.

[0063] Secondly, according to the search space of the control parameters, it is determined whether the current position of the particle exceeds the search space of the control parameters. If so, the current position of the particle is corrected to the boundary of the search space, otherwise, the current position of the particle is kept unchanged.

[0064] Determine whether the speed and position of all particles in the original particle swarm are updated. If so, perform the mutation operation to obtain the offspring particle swarm, calculate the fitness of each particle in the offspring particle swarm, and screen the second non-dominated particle of the offspring particle swarm. Otherwise, continue to update.

[0065] Specifically, the specific process of performing the mutation operation in step S26 is:

[0066] The original particle group is divided into three parts, the particles in the first part are kept not to perform mutation, the particles in the second part perform uniform mutation, and the particles in the third part perform non-uniform mutation.

[0067] Among them, the second non-dominated particle is a particle in the offspring particle swarm whose first objective function value and second objective function value are both less than or equal to the corresponding first objective function value and second objective function value of other particles, and is smaller in at least the first objective function value or the second objective function value.

[0068] S27. Insert the second non-dominated particle into the external particle swarm, compare the fitness of the first non-dominated particle and the second non-dominated particle, and obtain the third non-dominated particle of the external particle swarm, specifically:

[0069] If the first objective function value and the second objective function value of the first non-dominated particle are both less than or equal to the corresponding first objective function value and second objective function value of the second non-dominated particle, and is smaller in at least the first objective function value or the second objective function value, then retain the first non-dominated particle.

[0070] If the first objective function value and the second objective function value of the second non-dominated particle are both less than or equal to the corresponding first objective function value and second objective function value of the first non-dominated particle, and is smaller in at least the first objective function value or the second objective function value, then retain the second non-dominated particle.

[0071] Among them, the third non-dominated particle includes the retained first non-dominated particle and the second non-dominated particle.

[0072] S28. Determine whether the size of the external particle swarm exceeds the maximum allowable value. If so, calculate the crowding degree of each third non-dominated particle in the external particles and sort them in ascending order. Sequentially delete the third non-dominated particles with smaller crowding degrees from the external particle swarm until the size of the external particle swarm does not exceed the maximum allowable value. Otherwise, execute step S29.

[0073] S29. Determine whether the current position of the second non-dominated particle completely dominates its historical best position in the original particle swarm. If so, use the current position of the second non-dominated particle as the historical best position. Otherwise, determine whether the historical best position of the second non-dominated particle in the original particle swarm completely dominates its position in the offspring particle swarm. If so, keep the historical best position of the second non-dominated particle unchanged. Otherwise, neither of them dominates, and replace the historical best position of the second non-dominated particle in the original particle swarm with its current position with a random probability.

[0074] In this embodiment, complete domination means that the first objective function value and the second objective function value of a certain non-dominated particle at its current position are both less than the corresponding first objective function value and second objective function value at its historical best position.

[0075] Increment the current iteration count by 1, and use the offspring particle swarm as the new original particle swarm to execute step S25.

[0076] In this embodiment, according to the above optimization method, taking the rated operating condition of the unit (opening 70%, water head 195m) as an example, under the controller power control mode (PCM), the PID control parameters are optimized (referring to the actual power plant parameter settings, without considering the effect of the differential control parameter K D . The optimization results are as Figures 3 - 4 shown. The Pareto front of the control parameters is as Figure 3 shown. The sets of the two objective function values on the Pareto front are respectively normalized. The control parameters at three positions on the Pareto front (two endpoints A, C and the point B with the minimum Euclidean distance from the origin) are input into the system, and the corresponding time-domain response and damping characteristics are as Figure 4 shown. It can be seen from Figure 4 that point B takes into account both the time-domain response and the damping characteristics, and is the most suitable result.

[0077] S3. Determine whether the control parameter optimization has been completed for all operating conditions. If so, obtain the Pareto front under all operating conditions and execute step S6; otherwise, execute step S4.

[0078] S4. Obtain the opening and water head of the hydro-generator unit system operating under the current operating condition, obtain the turbine transfer coefficient under the current operating condition, and execute steps S1 - S2 to obtain and normalize the Pareto front of the hydro-generator unit system operating under the current operating condition.

[0079] In this embodiment, the purpose of normalization is to make the target values under different numerical scales all fall within the interval [0, 1].

[0080] S5. Calculate the Euclidean distance between each point on the normalized Pareto front and the origin, take the optimal solution with the minimum Euclidean distance as the optimal control parameter for the current operating condition, and at the same time increment the operating condition counter by 1, and execute step S3.

[0081] In this embodiment, each point on the normalized Pareto front refers to the position point of the optimal control parameter under each operating condition.

[0082] S6. Use the thin plate spline method to fit the Pareto front under all operating conditions to obtain the fitting surface of the optimal control parameter. By real-time detecting the current opening and water head of the hydro-generator unit system operating in the normal operating range, and based on the fitting surface, adjust and output the optimal value of the current control parameter.

[0083] Specifically, step S6 specifically includes S61 - S62:

[0084] S61. Use the thin plate spline method to fit the Pareto front under all operating conditions, with the water head and opening as the X-axis and Y-axis, and with the proportional adjustment coefficient K P or the integral adjustment coefficient K ITaking the Z-axis, a fitting surface for generating optimal control parameters is created.

[0085] S62. Real-time detect the current opening and water head when the hydro-generator unit system operates in the normal working range, and substitute the current opening and water head into the fitting surface of the optimal control parameters respectively to adjust and output the optimal value of the current control parameters in real time.

[0086] In this embodiment, although under a certain working condition, the method proposed in S1 - S2 can be used to obtain the set of optimal control parameters (i.e., the Pareto front) adapted to this working condition; however, due to the non-linearity of the water turbine and the fact that the operating conditions of the unit may change due to scheduling requirements, it is necessary to optimize the parameters of the controller for all operating conditions in the non-limited area of the unit, and then use the thin plate spline method to fit the data set of working conditions - optimal control parameters, and finally form an adaptive controller that automatically adjusts parameters according to the working conditions. Therefore, based on the above idea, a PID control strategy applicable to the multi-condition operation of the unit is proposed, specifically including the above steps S1 - S6. At the same time, according to the above steps, the spatial distribution of the optimal control parameters with the working conditions in the power control mode can be obtained, as Figures 5 - 6 shown, Figure 5 is the fitting surface for the proportional regulation coefficient K P regulation, Figure 6 is the fitting surface for the integral regulation coefficient K I regulation. Among them, the fitting is to take the water head and the opening as the x-axis and y-axis, and the control parameters of all working conditions as the z-axis to draw a three-dimensional graph.

[0087] In addition, in order to verify the effectiveness of a multi-condition adaptive control method for the optimal operation of a hydro-generator unit proposed in the present invention, by applying it to the speed regulation system of the hydro-generator unit, the superiority of the proposed method is proved. Specifically, the multi-objective condition adaptive controller generated by the method proposed in the present invention is compared with the controller optimized based on the ITAE (Integral of Time multiplied by the Absolute Error) index that only considers the primary frequency regulation performance in multiple scenarios. Specifically:

[0088] As Figure 7 shown, Figure 7 shows typical working conditions, that is, the water head and opening in the range of 155m - 225m and 60% - 100% in the normal operation range are divided into nine working conditions. The adaptive control results of four working conditions, namely LM, HM, ML, and MH, are selected and compared with the control results optimized based on the ITAE index that only considers the primary frequency regulation performance. Among them, the LM working condition is that the relative opening of the guide vane is 60% and the water head is 195m; the HM working condition is that the relative opening of the guide vane is 100% and the water head is 195m; the ML working condition is that the relative opening of the guide vane is 80% and the water head is 155m; the MH working condition is that the relative opening of the guide vane is 80% and the water head is 225m. The comparison results of the step disturbance under power control (PCM) are asFigures 8 - 11 As shown, the adaptive parameters corresponding to the four operating conditions of LM, HM, ML, and MH and the fixed-value optimization parameters considering only primary frequency regulation are respectively input into the system, and from Figures 8 - 11 It can be seen that when a disturbance occurs, the overshoot and offset of the adaptive controller under the four operating conditions are smaller than those of the controller considering only the primary frequency regulation performance, and the system can also return to stability faster. Therefore, the method proposed in the present invention is significantly superior to the ordinary control system method, and the method proposed in the present invention is applicable to the hydroturbine governing system with complex operating conditions. By adjusting the optimal control parameters according to the change of the operating conditions, the system can be kept stable and have good control performance.

[0089] In the present invention, specific embodiments are used to elaborate on the principle and implementation manner of the present invention. The description of the above embodiments is only used to help understand the method and its core idea of the present invention; at the same time, for those of ordinary skill in the art, according to the idea of the present invention, there will be changes in the specific implementation manner and application scope. In summary, the content of this specification should not be construed as a limitation to the present invention.

[0090] Those of ordinary skill in the art will realize that the embodiments described herein are for helping the reader understand the principle of the present invention, and it should be understood that the protection scope of the present invention is not limited to such specific statements and embodiments. Those of ordinary skill in the art can make various other specific deformations and combinations without departing from the essence of the present invention according to the technical revelations disclosed in the present invention, and these deformations and combinations are still within the protection scope of the present invention.

Claims

1. A multi-operating-condition adaptive control method for optimizing the operation of a hydropower unit, characterized in that: The following steps are involved: S1. Obtain several operating conditions that allow the hydropower unit system to operate, obtain the turbine transfer coefficient under a single operating condition, and calculate the state space equation of the hydropower unit system operation to determine the stable domain of the control parameters; S2. The stable domain of the control parameters is used as the search space of the control parameters. The particle fitness is calculated by introducing the comprehensive evaluation index of the control performance of the hydropower unit governor and the evaluation index of the unit damping characteristics, so as to optimize the control parameters of the hydropower unit system under each operating condition and obtain the Pareto frontier of the hydropower unit system under each operating condition. S3, judging whether the control parameter optimization is completed for all working conditions, if so, obtaining the Pareto frontier under all working conditions and executing step S6, otherwise, executing step S4; S4, obtaining the opening and water head of the hydropower unit system under the current working condition, obtaining the turbine transfer coefficient under the current working condition, and executing steps S1-S2 to obtain the Pareto front of the hydropower unit system under the current working condition and normalize it; S5, calculating the Euclidean distance between each point on the normalized Pareto front and the origin, taking the optimal solution with the minimum Euclidean distance as the optimal control parameter of the current working condition, and adding 1 to the working condition counter, and executing step S3; S6. Use the thin plate spline method to fit the Pareto front under all working conditions to obtain the fitting surface of the optimal control parameters. Through real-time detection of the current opening and water head of the hydropower unit system operating in the normal working range, and based on the fitting surface, adjust and output the optimal value of the current control parameter.

2. The multi-operating-condition adaptive control method for optimized operation of a hydropower unit according to claim 1 is characterized in that: The control parameter in step S1 is the proportional adjustment coefficient K P , integral adjustment coefficient K I .

3. The multi-operating-condition adaptive control method for optimized operation of a hydropower unit according to claim 2 is characterized in that: The state space equation of the hydropower unit system operation in step S1 is: Among them, Δdx′ represents the intermediate variable related to the differential, represents the first derivative of the intermediate variable related to the differential, Δix represents the integral variable, represents the first derivative of the integral variable, Δy represents the change in the servomotor stroke, represents the first derivative of the servomotor stroke change, Δh′ represents the intermediate variable related to the water head, represents the first-order derivative of the intermediate variable related to the water head, Δω represents the change in the angular velocity of the hydropower unit, Represents the first derivative of the angular velocity change of the hydropower unit, Δω r represents the given speed change, ω0 represents the initial angular velocity, E″ q represents the q-axis transient potential, E″ d Denotes the d-axis transient potential, ΔE″ q Indicates the change in the q-axis transient potential, ΔE″ d represents the change of the transient potential on the d-axis. The first derivative of the transient potential change on the d-axis, ΔE q ′ represents the change of transient potential on the q axis, The first-order derivative of the transient potential change on the q-axis, ΔE f represents the change of the excitation electromotive force, The first-order derivative of the change in the excitation electromotive force, K D Denotes the differential adjustment coefficient, K I Indicates the integral adjustment coefficient, K P ′ represents the intermediate variable of the proportional adjustment coefficient, K a represents the equivalent excitation system gain, δ represents the generator power angle, Δδ represents the change in the generator power angle, Represents the first-order derivative of the generator power angle change, X d represents the d-axis synchronous reactance, X′ d Indicates the d-axis transient reactance, X″ d represents the d-axis subtransient reactance, X q represents the q-axis synchronous reactance, X″ q represents the q-axis subtransient reactance, ΔP er Indicates the given active power change, T j Represents the inertia time constant of the hydropower unit, T d ′0 represents the d-axis open-circuit transient time constant, T d0 Indicates the d-axis open circuit subtransient time constant, T″ q0 represents the q-axis open-circuit subtransient time constant, T r Denotes equivalent time, T 1v represents the time constant of the differential link, T y represents the main relay reaction time constant, T w represents the water flow inertia time constant, e p Indicates the hydropower unit adjustment rate, e h represents the turbine torque to head transfer coefficient, e qy represents the transfer coefficient of turbine torque to guide vane opening, e qh represents the turbine flow to head transfer coefficient, e qx Indicates the turbine flow to speed transfer coefficient, ΔV gr Indicates the change in voltage at the given machine end, K 1,B , K 2,B , K 3,B , K 4,B , K 5,B , K 6,B , K 7,B , K 8,B , K 9,B , K 10,B , K 11,B , K 12,B , K 13,B They represent different intermediate variables respectively.

4. The multi-operating-condition adaptive control method for optimized operation of a hydropower unit according to claim 3 is characterized in that: The Pareto front in step S2 is the optimal control parameter set under multi-objective optimization.

5. The multi-operating-condition adaptive control method for optimized operation of a hydropower unit according to claim 4 is characterized in that: The comprehensive evaluation index of the control performance of the hydropower unit speed governor in step S2 is: Among them, H OF It represents the comprehensive evaluation index of the control performance of the hydropower unit speed governor. represents the absolute error time integral index, e(t) represents the error value between the expected output and the actual output of the hydropower unit system at the tth moment, F P1 represents the first penalty function, F P2 represents the second penalty function, if and else represent the judgment conditions, N B represents the penalty function value, T S It represents the adjustment time, that is, the shortest time required for the response curve to reach and no longer exceed the error band, T Smax Indicates the maximum allowable adjustment time, T R It represents the rise time, that is, the time required for the response to rise to the final value for the first time, T Rmin Indicates the shortest allowed rise time, A R Indicates the amount of reverse adjustment, A O Indicates the overshoot, A Rmax Indicates the maximum allowable reverse adjustment, A Omax Indicates the maximum allowable overshoot, A S It represents the disturbance amount, that is, the absolute value of the difference between the steady-state value before the disturbance and the steady-state value after the disturbance. C(t) represents the response curve. Min and max represent the minimum and maximum values, respectively.

6. The multi-operating-condition adaptive control method for optimized operation of a hydropower unit according to claim 5 is characterized in that: The evaluation index of the damping characteristic of the unit in step S2 is: Among them, D OF It represents the damping characteristic evaluation index of the hydropower unit, D s represents the overall damping value of the unit, f represents the oscillation frequency, D t (f) represents the unit damping corresponding to different oscillation frequencies f, real[·] represents the function to obtain the real part, s represents the Laplace operator, It represents the transfer function of the output turbine torque to the input speed, j represents the imaginary unit, ω d represents the angular velocity of oscillation.

7. The multi-operating-condition adaptive control method for optimized operation of a hydropower unit according to claim 6 is characterized in that: Step S2 specifically includes: S21, using the stable domain of the control parameter as the search space of the control parameter, setting the particle swarm size, the maximum number of iterations, the inertia weight, the learning factor and the external particle swarm size, initializing the particle swarm, and obtaining the original particle swarm; S22. Calculate the fitness of each particle in the original particle group, specifically: The first objective function and the second objective function are introduced, wherein the first objective function is a comprehensive evaluation index of the speed governor control performance of the hydropower unit, and the second objective function is an evaluation index of the damping characteristics of the unit, and the values ​​of the first objective function and the second objective function are used as the fitness of each particle in the original particle swarm; S23, taking the initial position of the particle in the original particle group as the historical optimal position of the particle; S24, inserting the first non-dominated particle in the original particle swarm into the external particle swarm, determining the grid where the first non-dominated particle in the external particle swarm is located, and counting the total number of the first non-dominated particles in each grid, calculating the probability of each grid being selected, and setting the current number of iterations to 1; The first non-dominated particle is a particle in the original particle group whose first objective function value and second objective function value are both less than or equal to the first objective function value and second objective function value corresponding to other particles, and at least the first objective function value or the second objective function value is smaller; S25, determine whether the current number of iterations reaches the maximum number of iterations, if so, end the search and output the optimal solution, otherwise, execute step S26; Among them, the optimal solution is the Pareto front; S26, based on the probability of each grid being selected, a roulette wheel method is used to select a grid, and a first non-dominated particle of the external particle swarm is randomly selected from the grid, the position of the first non-dominated particle is used as the global optimal position of the particle swarm, the position and speed of all particles in the original particle swarm are updated, and the speed and position of the particles are corrected according to the boundary constraints; Determine whether the speed and position of all particles in the original particle swarm have been updated. If so, perform the mutation operation to obtain the offspring particle swarm, calculate the fitness of each particle in the offspring particle swarm, and select the second non-dominated particle of the offspring particle swarm. Otherwise, continue to update; The specific process of performing mutation operation is as follows: The original particle group is divided into three parts, the particles in the first part are kept not to mutate, the particles in the second part are uniformly mutated, and the particles in the third part are non-uniformly mutated; The second non-dominated particle is a particle in the descendant particle group whose first objective function value and second objective function value are both less than or equal to the first objective function value and second objective function value corresponding to other particles, and at least the first objective function value or the second objective function value is smaller; S27, inserting the second non-dominated particle into the external particle group, and comparing the fitness of the first non-dominated particle with the second non-dominated particle, to obtain the third non-dominated particle of the external particle group, specifically: If the first objective function value and the second objective function value of the first non-dominated particle are both less than or equal to the first objective function value and the second objective function value corresponding to the second non-dominated particle, and at least the first objective function value or the second objective function value is smaller, then the first non-dominated particle is retained; If the first objective function value and the second objective function value of the second non-dominated particle are both less than or equal to the first objective function value and the second objective function value corresponding to the first non-dominated particle, and at least the first objective function value or the second objective function value is smaller, then the second non-dominated particle is retained; The third non-dominated particles include the retained first non-dominated particles and the second non-dominated particles; S28, determining whether the size of the external particle group exceeds the maximum allowed value, if so, calculating the crowding degree of each third non-dominated particle in the external particles, and sorting them in order from small to large, and sequentially deleting the third non-dominated particles with smaller crowding degrees from the external particle group until the size of the external particle group does not exceed the maximum allowed value, otherwise, executing step S29; S29, judging whether the current position of the second non-dominated particle completely dominates the historical optimal position of the second non-dominated particle in the original particle swarm, if so, taking the current position of the second non-dominated particle as the historical optimal position, otherwise, judging whether the historical optimal position of the second non-dominated particle in the original particle swarm completely dominates the position of the second non-dominated particle in the descendant particle swarm, if so, keeping the historical optimal position of the second non-dominated particle unchanged, otherwise, the two do not dominate each other, and replacing the historical optimal position of the second non-dominated particle in the original particle swarm with the current position of the second non-dominated particle with a random probability; The current iteration number is increased by 1, and the descendant particle group is used as the new original particle group to execute step S25.

8. The multi-operating-condition adaptive control method for optimized operation of a hydropower unit according to claim 7 is characterized in that: The formula for updating the positions and velocities of all particles in the original particle swarm in step S26 is: Among them, v i (t+1) represents the velocity of particle i at time t+1, v i (t) represents the velocity of particle i at time t, w represents the inertia weight, c1 and c2 represent acceleration constants, r1 and r2 represent random numbers between 0 and 1, and pbest i represents the historical optimal position of particle i, gbest represents the global optimal position of the particle swarm, and x i (t+1) represents the position of particle i at time t+1, x i (t) represents the position of particle i at time t.

9. The multi-operating-condition adaptive control method for optimized operation of a hydropower unit according to claim 8, characterized in that: The specific process of correcting the velocity and position of the particle according to the boundary constraint in step S26 is as follows: First, according to the set upper and lower limits of particle speed, determine whether the current speed of the particle is greater than the set maximum speed. If so, correct the current speed of the particle to the set maximum speed. Otherwise, determine whether the current speed of the particle is less than the set minimum speed. If so, correct the current speed of the particle to the set minimum speed. Otherwise, keep the current speed of the particle unchanged. Secondly, according to the search space of the control parameters, it is determined whether the current position of the particle exceeds the search space of the control parameters. If so, the current position of the particle is corrected to the boundary of the search space, otherwise, the current position of the particle is kept unchanged.

10. The multi-operating-condition adaptive control method for optimized operation of a hydropower unit according to claim 9, characterized in that: Step S6 specifically includes: S61, use the thin plate spline method to fit the Pareto front under all working conditions, with the water head and opening as the X-axis and Y-axis, and the proportional adjustment coefficient K P Or integral adjustment coefficient K I For the Z axis, generate the fitting surface of the optimal control parameters; S62, real-time detection of the current opening and water head of the hydropower unit system in the normal working range, and substitution of the current opening and water head into the fitting surface of the optimal control parameters, so as to adjust and output the optimal value of the current control parameter in real time.

Images