Coordinated control method for ultra-supercritical coal-fired power generation unit

Abstract

The invention provides a coordinated control method for an ultra-supercritical coal-fired power generating unit, and the method comprises the steps: constructing a sub-model set needed by a multi-model prediction control algorithm based on a gap measurement method, and independently carrying out the prediction and rolling optimization of each sub-controller through the prediction control algorithm; and the optimal control quantity is obtained by using the weighting strategy of the controller and acts on the unit coordination system. According to the invention, coordinated control of the ultra-supercritical coal-fired power generation unit can be realized, so that the unit can quickly track a load instruction and ensure stability of main steam pressure and outlet temperature of a steam-water separator, and the stability of the unit is improved; the fluctuation of fuel quantity is reduced, and the high efficiency of the unit is improved; the variable load rate of the unit is increased, and the flexibility of the unit is improved.

Description

Technical Field

[0001] The present invention relates to the field of thermal control technology, and in particular to a coordinated control method for an ultra-supercritical coal-fired power generation unit. Background Art

[0002] The strong volatility and lack of peak-shaving capabilities of renewable energy sources urgently require the support of deep peak-shaving capabilities of coal-fired power generation units. In coal-fired power generation units, the coordinated control system is the core of the entire automatic control system of modern coal-fired power plants, and it plays a key role in ensuring the safe and stable operation of power production. Because the dynamic characteristics of ultra-supercritical coal-fired power generation units exhibit strong nonlinear characteristics when operating at high parameters and wide loads, and the unit has severe multi-variable coupling under rapid load changes, key parameters such as main steam pressure, temperature, and unit load have serious mutual influence. This complex dynamic characteristic imposes many limitations on traditional PID control methods. Therefore, research on the coordinated control system of ultra-supercritical coal-fired power generation units using other advanced control methods is of great significance to ensuring the safety of ultra-supercritical coal-fired power generation units and increasing their economy and flexibility. Summary of the Invention

[0003] In order to overcome the deficiencies of the prior art, the present invention aims to provide a coordinated control method for an ultra-supercritical coal-fired power generation unit, which can improve the stability, efficiency and flexibility of the ultra-supercritical coal-fired power generation unit.

[0004] In order to achieve the above object, the present invention adopts the following technical solutions:

[0005] A coordinated control method for an ultra-supercritical coal-fired power generation unit comprises the following steps:

[0006] Step S1: constructing a nonlinear model of the coordinated system of ultra-supercritical coal-fired power generation units;

[0007] Step S2: constructing a sub-model set of the nonlinear model of the ultra-supercritical coal-fired power generation unit coordination system required by the multi-model predictive control algorithm based on the gap measurement method, and designing corresponding sub-controllers using the sub-model set as the prediction model in the multi-model predictive control algorithm;

[0008] Step S3: After the sub-controller obtains the unit load command value, the main steam pressure set value, and the steam-water separator outlet temperature set value signals, each sub-controller performs independent prediction based on the predictive control algorithm to obtain the control sequence in the future time domain;

[0009] Step S4: Each sub-controller performs online rolling optimization to solve the optimal control sequence that meets the constraints, and inputs the optimal control sequence into the sub-model set to obtain the predicted output value of the prediction model;

[0010] Step S5: performing weighting according to the weighting strategy and applying the final control variable to the nonlinear model of the unit coordination system constructed in step S1 to obtain the actual output value of the nonlinear model of the unit coordination system;

[0011] Step S6: comparing the actual output value of the nonlinear model of the unit coordination system with the predicted output value of the prediction model, calculating the error, and performing feedback correction on the reference value input to the sub-controller;

[0012] Step S7: Roll the optimization time domain to perform prediction and optimization for the next cycle, return to step S3 and repeat this process.

[0013] In step S1, the nonlinear model of the ultra-supercritical coal-fired power generation unit coordination system is a three-input three-output model, wherein the input variables are the coal feed rate, the feed water flow rate and the turbine throttle opening; and the output variables are the unit load, the main steam pressure and the steam-water separator outlet temperature.

[0014] In step S2, the steps of constructing a sub-model set of the nonlinear model of the ultra-supercritical coal-fired power generation unit coordination system required by the multi-model predictive control algorithm based on the gap measurement method are as follows:

[0015] Operating range division: Based on the operating load range of the ultra-supercritical coal-fired power generation unit, it is divided into equally spaced sub-operating points according to the unit load points;

[0016] Model linearization: linearize each sub-operating point to obtain the corresponding sub-model;

[0017] Gap metric calculation: Calculate the gap metric between each two sub-models;

[0018] Optimization and adjustment of operating points: Starting from the sub-model with the lowest load point, if the gap measurement value of adjacent sub-models is less than the set threshold, the adjacent sub-models are merged to reduce redundancy; if the gap measurement value of adjacent sub-models is greater than the set threshold, a new operating point is inserted into it using the dichotomy method to enable it to express the dynamic characteristics between the operating points;

[0019] Model set reconstruction: After optimizing and adjusting the operating points, a sub-model set covering all operating conditions is obtained.

[0020] The linearization process for each sub-operating point is:

[0021]

[0022] Among them, P ab is the transfer function of the bth input variable to the ath output variable; Pe is the unit load, MW; p st Main steam pressure, MPa; h sep is the separator outlet enthalpy, °C; uB is the coal feeding rate, kg / s; D fw is the water flow rate, kg / s; μ T Adjust the valve opening for the steam turbine.

[0023] In step S4, the performance indicator function during the online rolling optimization process is:

[0024]

[0025] Where J is the performance index function; w(k+j) is the expected output value at time k+j; y(k+j|k) is the predicted output value at time k+j; u(k+j|k) is the control input value at time k+j; Q and R are the error weight matrices of output and control quantities;

[0026] The optimal control sequence is determined by finding the minimum value of the performance indicator function.

[0027] In step S4, the constraints include the control amount amplitude constraint, the control amount change rate amplitude constraint, and the output amount amplitude constraint:

[0028] u min ≤u(k+j|k)≤u max

[0029] Δu min ≤u(k+j|k)-u(k+j-1|k)≤Δu max

[0030] y min ≤y(k+j|k)≤y max

[0031] Among them, u max 、u min is the upper and lower limits of the control amplitude; Δu max , Δu min is the upper and lower limits of the control variable change rate; y max 、y min ——Upper and lower limits of output amplitude.

[0032] In step S5, the process of weighting according to the weighting strategy is as follows:

[0033] The weighting strategy adopts the improved recursive Bayesian weighting algorithm:

[0034]

[0035] Among them, d i,k is the conditional probability of the matching degree between the i-th sub-model and the linear model corresponding to the current working condition; δ i,kis the gap measurement value between the i-th sub-model and the linear model corresponding to the current working condition; E is the convergence coefficient; n1 is the number of sub-models in the newly constructed sub-model set; i,k Normalize to get the weight value corresponding to each sub-model:

[0036]

[0037] ω i is the weight value of the i-th sub-model.

[0038] The beneficial effects of the technical solution provided by the present invention include at least:

[0039] The present invention provides a coordinated control method for ultra-supercritical coal-fired power generation units. This coordinated control method constructs the sub-model set required by a multi-model predictive control algorithm based on a gap metric method. Each sub-controller independently performs prediction and rolling optimization using the predictive control algorithm. A weighted strategy for the sub-controllers is used to determine the optimal control variable, which is then applied to the unit coordination system. The present invention enables coordinated control of ultra-supercritical coal-fired power generation units, enabling the units to rapidly track load commands and maintain the stability of the main steam pressure and steam-water separator outlet temperature, thereby improving unit stability; reducing fuel quantity fluctuations, thereby increasing unit efficiency; and increasing the unit's load-variable rate, thereby enhancing unit flexibility. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 The present invention is a flow chart of a coordinated control method for an ultra-supercritical coal-fired power generation unit;

[0041] Figure 2 This is a diagram of the coordinated control system of the ultra-supercritical coal-fired power generation unit in the present invention;

[0042] FIG3 is a graph showing the variation of model output variables during the simulation test of the present invention, wherein (a) is a curve showing the variation of the model output variable unit load over time, (b) is a curve showing the variation of the model output variable main steam pressure over time, and (c) is a curve showing the variation of the model output variable steam-water separator outlet temperature over time;

[0043] Figure 4 This is a fuel quantity variation curve during the simulation test of the present invention. DETAILED DESCRIPTION

[0044] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0045] like Figure 1 As shown, the present invention is a coordinated control method for an ultra-supercritical coal-fired power generation unit, comprising the following steps:

[0046] A coordinated control method for an ultra-supercritical coal-fired power generation unit comprises the following steps:

[0047] Step S1: constructing a nonlinear model of the coordinated system of ultra-supercritical coal-fired power generation units;

[0048] The nonlinear model of the coordinated system of ultra-supercritical coal-fired power generation units is a three-input and three-output model, wherein the input variables are coal feed rate, feed water flow rate and turbine throttle opening; and the output variables are unit load, main steam pressure and steam-water separator outlet temperature.

[0049] Step S2: constructing a sub-model set of the nonlinear model of the ultra-supercritical coal-fired power generation unit coordination system required by the multi-model predictive control algorithm based on the gap measurement method, and designing corresponding sub-controllers using the sub-model set as the prediction model in the multi-model predictive control algorithm;

[0050] Specifically, the steps for constructing the sub-model set of the nonlinear model of the ultra-supercritical coal-fired power generation unit coordination system required by the multi-model predictive control algorithm based on the gap measurement method are as follows:

[0051] Operating range division: Based on the operating load range of the ultra-supercritical coal-fired power generation unit, it is divided into equally spaced sub-operating points according to the unit load points;

[0052] Model linearization: linearize each sub-operating point to obtain the corresponding sub-model;

[0053] Gap metric calculation: Calculate the gap metric between each two sub-models;

[0054] Optimization and adjustment of operating points: Starting from the sub-model with the lowest load point, if the gap measurement value of adjacent sub-models is less than the set threshold, the adjacent sub-models are merged to reduce redundancy; if the gap measurement value of adjacent sub-models is greater than the set threshold, a new operating point is inserted into it using the dichotomy method to enable it to express the dynamic characteristics between the operating points;

[0055] Model set reconstruction: After optimizing and adjusting the operating points, a sub-model set covering all operating conditions is obtained.

[0056] Specifically, the linearization process for each sub-operating point is:

[0057]

[0058] Among them, P ab is the transfer function of the bth input variable to the ath output variable; Pe is the unit load, MW; p st Main steam pressure, MPa; h sep is the separator outlet enthalpy, °C; u B is the coal feeding rate, kg / s; D fw is the water flow rate, kg / s; μ T Adjust the valve opening for the steam turbine.

[0059] Step S3: After the sub-controller obtains the unit load command value, the main steam pressure set value, and the steam-water separator outlet temperature set value signals, each sub-controller performs independent prediction based on the predictive control algorithm to obtain the control sequence in the future time domain;

[0060] Step S4: Each sub-controller performs online rolling optimization to solve the optimal control sequence that meets the constraints, and inputs the optimal control sequence into the sub-model set to obtain the predicted output value of the prediction model;

[0061] Specifically, the performance indicator function during online rolling optimization is:

[0062]

[0063] Where J is the performance index function; w(k+j) is the expected output value at time k+j; y(k+j|k) is the predicted output value at time k+j; u(k+j|k) is the control input value at time k+j; Q and R are the error weight matrices of output and control quantities;

[0064] The optimal control sequence is determined by finding the minimum value of the performance indicator function.

[0065] Specifically, the constraints include the control quantity amplitude constraint, the control quantity change rate amplitude constraint, and the output quantity amplitude constraint:

[0066] u min ≤u(k+j|k)≤u max

[0067] Δu min ≤u(k+j|k)-u(k+j-1|k)≤Δu max

[0068] y min ≤y(k+j|k)≤y max

[0069] Among them, u max 、u min is the upper and lower limits of the control amplitude; Δu max , Δu min is the upper and lower limits of the control variable change rate; y max 、y min ——Upper and lower limits of output amplitude.

[0070] Step S5: performing weighting according to the weighting strategy and applying the final control variable to the nonlinear model of the unit coordination system constructed in step S1 to obtain the actual output value of the nonlinear model of the unit coordination system;

[0071] In step S5, the process of weighting according to the weighting strategy is as follows:

[0072] The weighting strategy adopts the improved recursive Bayesian weighting algorithm:

[0073]

[0074] Among them, d i,k is the conditional probability of the matching degree between the i-th sub-model and the linear model corresponding to the current working condition; δ i,k is the gap measurement value between the i-th sub-model and the linear model corresponding to the current working condition; E is the convergence coefficient; n1 is the number of sub-models in the newly constructed sub-model set; i,k Normalize to get the weight value corresponding to each sub-model:

[0075]

[0076] ω i is the weight value of the i-th sub-model.

[0077] Step S6: comparing the actual output value of the nonlinear model of the unit coordination system with the predicted output value of the prediction model, calculating the error, and performing feedback correction on the reference value input to the sub-controller;

[0078] Step S7: Roll the optimization time domain to perform prediction and optimization for the next cycle, return to step S3 and repeat this process.

[0079] The following describes the content of the present invention by taking a domestic 650MW ultra-supercritical coal-fired power generation unit adopting the coordinated control method of the present invention as an example.

[0080] In this example, a nonlinear model for the coordinated system of ultra-supercritical coal-fired power generation units (USCs) was developed using a gray-box modeling approach based on mechanism analysis and data-driven modeling, taking into account the nonlinear characteristics of the coordinated system, including multiple inputs, multiple outputs, and strong coupling. The model's state variables are fuel quantity, separator outlet pressure, and separator outlet enthalpy; the input variables are coal feed rate, feedwater flow rate, and turbine throttle opening; and the output variables are unit load, main steam pressure, and separator outlet temperature.

[0081] The sub-model set constructed based on the gap measurement method includes four sub-models, corresponding to the operating points of 300MW, 375MW, 470MW, and 565MW. The sub-models are (taking the sub-model at the 300MW load point as an example):

[0082]

[0083] in,

[0084]

[0085] After the sub-model set is used as a prediction model to design the corresponding sub-controller, the prediction and rolling optimization process are carried out based on the predictive control algorithm, and weighting and feedback correction are performed according to the weighting strategy, and finally the coordinated control system loop of the ultra-supercritical coal-fired power generation unit is formed, such as Figure 2 shown.

[0086] For the unit, the temperature is increased from 50%THA to 1.0%Pe0·min. -1 A simulation experiment was conducted using a variable load process where the load rate rises to 75% THA and compared with a traditional PID control method. The curves of the model output variables during this process are shown in Figures 3 (a), (b), and (c). The corresponding data are shown in Tables 1 and 2 below:

[0087] Table 1

[0088]

[0089] Table 2

[0090]

[0091] As can be seen from Figure 3 and the table, compared with the traditional PID control method, the control method of the present invention significantly reduces the adjustment time and overshoot of the unit output variable. The unit can quickly track the load command and ensure the stability of the main steam pressure and the steam-water separator outlet temperature, thereby improving the stability of the unit.

[0092] The fuel quantity change curve during this process is shown in the figure below: Figure 4 As shown in the figure, it can be seen that the control method of the present invention reduces the fluctuation of the fuel amount and improves the efficiency of the unit compared with the traditional PID control method.

[0093] During this process, the control method of the present invention can achieve a maximum load change rate of 2.0% Pe0·min within the safety threshold range. -1 Compared with the traditional PID control method, it improves the load change rate and the flexibility of the unit.

[0094] The present invention's coordinated control method for ultra-supercritical coal-fired power generation units constructs the sub-model set required by the multi-model predictive control algorithm based on a gap metric method. Each sub-controller independently performs prediction and rolling optimization using the predictive control algorithm. The controller's weighted strategy is used to obtain the optimal control variable, which is then applied to the unit's coordinated system. The present invention enables coordinated control of ultra-supercritical coal-fired power generation units, enabling the units to quickly track load commands and ensure the stability of the main steam pressure and steam-water separator outlet temperature, thereby improving the unit's stability; reducing fuel quantity fluctuations, thereby increasing the unit's efficiency; and increasing the unit's load-variable rate, thereby enhancing the unit's flexibility.

Claims

1. A coordinated control method for an ultra-supercritical coal-fired power generation unit, characterized in that: The steps include: Step S1: constructing a nonlinear model of the coordinated system of ultra-supercritical coal-fired power generation units; Step S2: constructing a sub-model set of the nonlinear model of the ultra-supercritical coal-fired power generation unit coordination system required by the multi-model predictive control algorithm based on the gap measurement method, and designing corresponding sub-controllers using the sub-model set as the prediction model in the multi-model predictive control algorithm; Step S3: After the sub-controller obtains the unit load command value, the main steam pressure set value, and the steam-water separator outlet temperature set value signal, each sub-controller performs independent prediction based on the predictive control algorithm () to obtain the control sequence in the future time domain; Step S4: Each sub-controller performs online rolling optimization to solve the optimal control sequence that meets the constraints, and inputs the optimal control sequence into the sub-model set to obtain the predicted output value of the prediction model; Step S5: performing weighting according to the weighting strategy and applying the final control variable to the nonlinear model of the unit coordination system constructed in step S1 to obtain the actual output value of the nonlinear model of the unit coordination system; Step S6: comparing the actual output value of the nonlinear model of the unit coordination system with the predicted output value of the prediction model, calculating the error, and performing feedback correction on the reference value input to the sub-controller; Step S7: Roll the optimization time domain to perform prediction and optimization for the next cycle, return to step S3 and repeat this process.

2. The coordinated control method of an ultra-supercritical coal-fired power generation unit according to claim 1, characterized in that: In step S1, the nonlinear model of the ultra-supercritical coal-fired power generation unit coordination system is a three-input three-output model, wherein the input variables are the coal feed rate, the feed water flow rate and the turbine throttle opening; and the output variables are the unit load, the main steam pressure and the steam-water separator outlet temperature.

3. The coordinated control method of an ultra-supercritical coal-fired power generation unit according to claim 1, characterized in that: In step S2, the steps of constructing a sub-model set of the nonlinear model of the ultra-supercritical coal-fired power generation unit coordination system required by the multi-model predictive control algorithm based on the gap measurement method are as follows: Operating range division: Based on the operating load range of the ultra-supercritical coal-fired power generation unit, it is divided into equally spaced sub-operating points according to the unit load points; Model linearization: linearize each sub-operating point to obtain the corresponding sub-model; Gap metric calculation: Calculate the gap metric between each two sub-models; Optimization and adjustment of operating points: Starting from the sub-model with the lowest load point, if the gap measurement value of adjacent sub-models is less than the set threshold, the adjacent sub-models are merged to reduce redundancy; if the gap measurement value of adjacent sub-models is greater than the set threshold, a new operating point is inserted into it using the dichotomy method to enable it to express the dynamic characteristics between the operating points; Model set reconstruction: After optimizing and adjusting the operating points, a sub-model set covering all operating conditions is obtained.

4. The coordinated control method for an ultra-supercritical coal-fired power generation unit according to claim 3, characterized in that: The linearization process for each sub-operating point is: Among them, P ab is the transfer function of the bth input variable to the ath output variable; Pe is the unit load, MW; p st Main steam pressure, MPa; h sep is the separator outlet enthalpy, °C; u B is the coal feeding rate, kg / s; D fw is the water flow rate, kg / s; μ T Adjust the valve opening for the steam turbine.

5. The coordinated control method of an ultra-supercritical coal-fired power generation unit according to claim 1, characterized in that: In step S4, the performance indicator function during the online rolling optimization process is: Where J is the performance index function; w(k+j) is the expected output value at time k+j; y(k+j|k) is the predicted output value at time k+j; u(k+j|k) is the control input value at time k+j; Q and R are the error weight matrices of output and control quantities; The optimal control sequence is determined by finding the minimum value of the performance indicator function.

6. The coordinated control method of an ultra-supercritical coal-fired power generation unit according to claim 1, characterized in that: In step S4, the constraints include the control amount amplitude constraint, the control amount change rate amplitude constraint, and the output amount amplitude constraint: u min ≤u(k+j|k)≤u max Δu min ≤u(k+j|k)-u(k+j-1|k)≤Δu max y min ≤y(k+j|k)≤y max Among them, u max 、u min is the upper and lower limits of the control amplitude; Δu max , Δu min is the upper and lower limits of the control variable change rate; y max 、y min ——Upper and lower limits of output amplitude.

7. The coordinated control method of an ultra-supercritical coal-fired power generation unit according to claim 1, characterized in that: In step S5, the process of weighting according to the weighting strategy is as follows: The weighting strategy adopts the improved recursive Bayesian weighting algorithm: Among them, d i,k is the conditional probability of the matching degree between the i-th sub-model and the linear model corresponding to the current working condition; δ i,k is the gap measurement value between the i-th sub-model and the linear model corresponding to the current working condition; E is the convergence coefficient; n1 is the number of sub-models in the newly constructed sub-model set; i,k Normalize to get the weight value corresponding to each sub-model: ω i is the weight value of the i-th sub-model.

Images