A method for linkage regulation of household garbage grate combustion and waste heat boiler load

Abstract

The application discloses a household garbage grate combustion and waste heat boiler load linkage regulation method, and relates to the automatic control field, which comprises the following steps: 1, collecting the operation data of the grate combustion side and the waste heat boiler side, setting an upper model predictive controller to rollingly predict and rollingly optimize the steam parameters, generating a steam load setting track and decomposing the steam load setting track into a hearth heat release setting value, a feed water setting value and a superheated steam temperature setting value; 2, setting an expansion state observer and a self-disturbance control controller on the grate combustion side, merging the household garbage calorific value and water content fluctuation and other disturbances into total disturbances, taking the total disturbance estimation value as an equivalent fuel disturbance amount, and adjusting the amount of household garbage fed into the furnace according to the equivalent fuel disturbance amount; 3, taking the equivalent fuel disturbance amount as a shared feedforward amount and simultaneously sending the equivalent fuel disturbance amount to a feed water control loop and a superheated steam temperature control loop, and adjusting the feed water and the desuperheating water flow in advance before the hearth heat release change is transmitted to the steam parameters. The application can stabilize the steam parameters without measuring the calorific value.

Description

Technical Field

[0001] This invention relates to the field of automatic control technology, specifically to a method for linking the combustion of municipal solid waste incinerator grate with the load of waste heat boiler, belonging to the field of industrial control software. Background Technology

[0002] Municipal solid waste incineration power generation is a major method for reducing, harmlessly treating, and recycling municipal solid waste. Its core equipment consists of a mechanical grate incinerator and a waste heat boiler. Municipal solid waste undergoes drying, ignition, combustion, and burnout on the grate, and the released heat is converted into steam by the heat exchanger in the waste heat boiler for power generation. In actual operation, there is a large inertia and long time lag between grate incineration and steam production in the waste heat boiler. From the time municipal solid waste is pushed into the furnace to the time the corresponding heat is reflected in the steam parameters, it often takes tens of seconds or even several minutes, causing fluctuations in steam flow rate, steam pressure, and superheated steam temperature. Currently, incineration lines generally adopt automatic combustion control based on proportional-integral-derivative (PID) logic. The furnace temperature and flue gas oxygen content are maintained through loop adjustments of grate speed, primary air volume, and secondary air volume. On the waste heat boiler side, the steam-water balance and superheated steam temperature are maintained by the feedwater loop and the desuperheating water loop, respectively. To cope with the long time lag, some incineration lines also introduce model predictive control to perform rolling optimization of the steam load. However, municipal solid waste is highly mixed in composition, with its calorific value and moisture content fluctuating randomly and unable to be directly measured before entering the furnace, constituting the primary source of disturbance in the combustion process. Proportional-integral-derivative (PID) control adjusts based on existing deviations, always resulting in a lag in suppressing this disturbance. Furthermore, the controls on the grate combustion side and the waste heat boiler side are largely independent, with the waste heat boiler side only passively adjusting after steam parameters deviate, making coordination with the combustion side difficult. Therefore, the problem of significant steam parameter fluctuations caused by fluctuations in the calorific value of municipal solid waste persists in existing technologies. Summary of the Invention

[0003] To address the problems in existing technologies, such as the inability to directly measure the calorific value and moisture content of municipal solid waste, significant fluctuations in steam parameters caused by calorific value fluctuations, and the lack of coordination between the grate combustion side and the waste heat boiler side, this invention provides a method for coordinated regulation of grate combustion and waste heat boiler load in municipal solid waste incinerators. This method unifies the fluctuations in calorific value and moisture content of municipal solid waste, along with other disturbances during grate combustion, into a total disturbance acting on the grate combustion process. An extended state observer estimates this total disturbance online and outputs it as an equivalent fuel disturbance. This same equivalent fuel disturbance is then used as a shared feedforward quantity, simultaneously provided to both the grate combustion side and the waste heat boiler side. This allows the grate combustion side to adjust the amount of municipal solid waste fed into the furnace, and the waste heat boiler side to adjust the feedwater and desuperheating water flow rates in advance. Thus, without the need for direct measurement of the calorific value of municipal solid waste, steam parameters remain stable despite fluctuations in the calorific value of the waste.

[0004] To achieve the above objectives, the technical solution adopted by the present invention is as follows: A method for linking the combustion of municipal solid waste grate with the load of a waste heat boiler for continuous operation, including municipal solid waste receiving, storage, feeding, grate combustion, and heat absorption and steam generation by the waste heat boiler heating surface, includes the following steps: Step 1: Real-time acquisition of operating data from the grate combustion side and the waste heat boiler side; setting up an upper-level model predictive controller, which uses the thermal inertia relationship between grate combustion and waste heat boiler as the predictive model to perform rolling prediction and rolling optimization of steam parameters, generate steam load setting trajectory, and decompose the steam load setting trajectory into furnace heat release setting value sent to the grate combustion side, feedwater setting value sent to the waste heat boiler side, and superheated steam temperature setting value sent to the waste heat boiler side. Step 2: An expansion state observer and an active disturbance rejection controller are installed on the grate combustion side. The expansion state observer takes the grate speed execution command, the primary air volume execution command, and the furnace temperature as inputs, and outputs the furnace heat release estimate and the total disturbance estimate. The expansion state observer unifies the fluctuations in the calorific value of municipal solid waste, the fluctuations in the moisture content of municipal solid waste, and other disturbances during the grate combustion process into a total disturbance acting on the grate combustion process, and uses the total disturbance estimate as an equivalent fuel disturbance. The active disturbance rejection controller forms the execution quantity on the grate combustion side based on the furnace heat release setpoint, the furnace heat release estimate, and the equivalent fuel disturbance. The execution quantity generates a grate speed execution command to adjust the amount of municipal solid waste fed into the furnace, so that when the equivalent fuel disturbance increases, the amount of municipal solid waste fed into the furnace decreases, and when the equivalent fuel disturbance decreases, the amount of municipal solid waste fed into the furnace increases. Step 3: The equivalent fuel disturbance output from the expanded state observer is used as a shared feedforward quantity and simultaneously sent to the feedwater control loop and superheated steam temperature control loop on the waste heat boiler side. The feedwater control loop adjusts the feedwater flow rate according to the feedwater setpoint and the equivalent fuel disturbance quantity, and the superheated steam temperature control loop adjusts the desuperheating water flow rate according to the superheated steam temperature setpoint and the equivalent fuel disturbance quantity. Before the change in furnace heat release is transmitted to the steam parameters, the waste heat boiler side adjusts the feedwater flow rate and desuperheating water flow rate in advance according to the equivalent fuel disturbance quantity. The grate combustion side and the waste heat boiler side use the same equivalent fuel disturbance quantity as a shared link to keep the steam parameters stable when the calorific value of municipal solid waste fluctuates, thereby realizing the linkage regulation of municipal solid waste incinerator grate combustion and waste heat boiler load.

[0005] Furthermore, the real-time operational data collected in step 1 includes the weighing of municipal solid waste entering the plant, the storage capacity of the waste bin zones, the amount of waste grabbed by the garbage crane, the frequency of municipal solid waste feeding, the grate speed, the primary air volume, the secondary air volume, the furnace temperature, the oxygen content of the flue gas, the carbon monoxide concentration, the steam flow rate, the steam pressure, the superheated steam temperature, and the ash status of the waste heat boiler heating surface. The weighing of municipal solid waste entering the plant, the storage capacity of the waste bin zones, and the amount of waste grabbed by the garbage crane represent the supply of municipal solid waste into the furnace. The furnace temperature, the oxygen content of the flue gas, and the carbon monoxide concentration represent the combustion of municipal solid waste on the grate. The steam flow rate, the steam pressure, and the superheated steam temperature represent the steam production of the waste heat boiler. The amount of municipal solid waste entering the furnace is jointly determined by the grate speed and the frequency of municipal solid waste feeding. The grate speed execution command changes the amount of municipal solid waste pushed into the furnace for incineration per unit time by adjusting the grate speed.

[0006] Furthermore, the steam parameters mentioned in step 1 include steam flow rate, steam pressure, and superheated steam temperature; the thermal inertia relationship characterizes the delay and correspondence between changes in furnace heat release and the thermal inertia of the waste heat boiler, which are transmitted to steam flow rate, steam pressure, and superheated steam temperature; in each scheduling cycle, the upper-level model predictive controller performs rolling predictions of steam flow rate, steam pressure, and superheated steam temperature for subsequent set periods based on the prediction model, and performs rolling optimization with the goal of the steam flow rate, steam pressure, and superheated steam temperature obtained from the rolling prediction approaching the steam load demand, generating a steam load setting trajectory to maintain the steam pressure within the set range; based on the energy correspondence between steam production from the waste heat boiler and furnace heat release, the upper-level model predictive controller decomposes the steam load setting trajectory into furnace heat release setpoints, feedwater setpoints, and superheated steam temperature setpoints, all of which are target values ​​given by the optimization layer.

[0007] Furthermore, in step 2, the extended state observer internally maintains three estimates: the furnace heat release estimate, the furnace heat release rate estimate, and the total disturbance estimate. All three estimates are observed quantities. The nominal input-output correspondence of the grate combustion process characterizes the influence of grate speed and primary air volume on furnace temperature under nominal municipal solid waste quality. In each control cycle, the extended state observer predicts the furnace temperature based on the current grate speed execution command, primary air volume execution command, and the nominal input-output correspondence of the grate combustion process. The predicted furnace temperature is compared with the furnace temperature collected in the next control cycle to obtain the observation deviation. Based on the observation deviation, the furnace heat release estimate, the furnace heat release rate estimate, and the total disturbance estimate are simultaneously corrected, so that the observation deviation converges as the control cycle progresses, and the furnace heat release estimate and the total disturbance estimate approach the actual value as the control cycle progresses.

[0008] Furthermore, in step 2, the calorific value and moisture content that change with the quality of municipal solid waste are estimated online by the expansion state observer as quantities to be estimated; the equivalent fuel disturbance represents the deviation trend of the actual energy released by the municipal solid waste entering the furnace per unit time relative to the energy required to maintain the furnace heat release setpoint.

[0009] Furthermore, in step 2, a tracking differentiator is set within the active disturbance rejection controller (ADRC). The tracking differentiator arranges a transition process for the furnace heat release setpoint, outputting the furnace heat release transition setpoint and its changing trend, thus transforming the step change in the furnace heat release setpoint into a smooth transition setpoint. The ADRC uses the deviation between the furnace heat release transition setpoint and the furnace heat release estimate as the tracking deviation. The tracking deviation and its changing trend form a feedback adjustment quantity, which increases with the increase of the tracking deviation and the increase of the rate of change of the tracking deviation. The ADRC converts the equivalent fuel disturbance into a feedforward compensation quantity, which is used to offset the effect of the equivalent fuel disturbance on the furnace heat release. The ADRC then superimposes the feedback adjustment quantity and the feedforward compensation quantity to obtain the execution quantity on the grate combustion side.

[0010] Furthermore, in step 2, the grate speed execution command drives the grate hydraulic drive mechanism to adjust the grate speed to adjust the amount of domestic waste entering the furnace; the primary air volume execution command changes in the same direction as the grate speed execution command and is corrected according to the oxygen content of the flue gas. The primary air volume is adjusted by the primary air regulating damper to maintain the ratio of primary air volume to the amount of domestic waste entering the furnace; the secondary air volume execution command is formed based on the oxygen content and carbon monoxide concentration of the flue gas. The secondary air volume is adjusted by the secondary air regulating damper to maintain the complete combustion of domestic waste on the grate.

[0011] Furthermore, in step 3, both the feedwater control loop and the superheated steam temperature control loop are control loops that combine feedforward and feedback. The feedwater control loop uses the collected steam flow rate and steam pressure as feedback quantities, and forms feedwater feedback regulation based on the deviation between the feedwater setpoint and the feedback quantity. The feedwater feedback regulation suppresses the deviation caused by the lag on the waste heat boiler side water side and changes in steam demand. The superheated steam temperature control loop uses the collected superheated steam temperature as feedback quantity, and forms desuperheating water feedback regulation based on the deviation between the superheated steam temperature setpoint and the feedback quantity. The desuperheating water feedback regulation suppresses the superheated steam temperature deviation caused by changes in the ash state of the waste heat boiler heating surface.

[0012] Furthermore, in step 3, the equivalent fuel disturbance is simultaneously sent to both the feedwater control loop and the superheated steam temperature control loop via the signal channel in each control cycle, ensuring that both loops receive the same equivalent fuel disturbance value within the same control cycle. Based on its own response characteristics to changes in feedwater flow, the feedwater control loop converts the received equivalent fuel disturbance into a feedwater feedforward correction. This correction increases with increasing equivalent fuel disturbance and decreases with decreasing equivalent fuel disturbance. The feedwater control loop then superimposes the feedwater feedforward correction onto the feedwater feedback regulation and adjusts the control based on the superimposed result. The feedwater flow rate is adjusted accordingly. The superheated steam temperature control loop, based on its own response characteristics to changes in the desuperheating water flow rate, converts the received equivalent fuel disturbance into a desuperheating water feedforward correction. The desuperheating water feedforward correction increases with the increase of the equivalent fuel disturbance and decreases with the decrease of the equivalent fuel disturbance. The superheated steam temperature control loop superimposes the desuperheating water feedforward correction on the basis of the desuperheating water feedback regulation and adjusts the desuperheating water flow rate based on the superimposed result. The feedwater feedforward correction and the desuperheating water feedforward correction are formed independently from equivalent fuel disturbances of the same value, so that the feedwater flow rate and the desuperheating water flow rate are corrected in advance according to the response characteristics of their respective loops.

[0013] Furthermore, in step 3, before the change in furnace heat release is transferred to the steam flow, steam pressure, and superheated steam temperature through the combustion of domestic waste on the grate and the thermal inertia of the waste heat boiler, the waste heat boiler side adjusts the feedwater flow and desuperheating water flow in advance based on the equivalent fuel disturbance, so that the heat absorption capacity of the waste heat boiler side is preset before the generated or reduced steam heat reaches the heating surface; when the furnace heat release increases, the grate combustion side changes the amount of domestic waste entering the furnace in a decreasing direction to bring the furnace heat release back to the furnace heat release set value, and the waste heat boiler side simultaneously increases the feedwater flow and desuperheating water flow to absorb the extra steam heat generated during the furnace heat release increase phase; when the furnace heat release decreases, the grate combustion side changes the amount of domestic waste entering the furnace in a increasing direction to bring the furnace heat release back to the furnace heat release set value, and the waste heat boiler side simultaneously decreases the feedwater flow and desuperheating water flow.

[0014] Compared with the prior art, the beneficial effects of the present invention are as follows: First, this invention unifies the fluctuations in calorific value and moisture content of municipal solid waste, as well as other disturbances during the grate combustion process, into a total disturbance acting on the grate combustion process. The equivalent fuel disturbance is estimated online by the expansion state observer based solely on the grate speed execution command, the primary air volume execution command, and the furnace temperature. This eliminates the need to directly measure the calorific value and moisture content of municipal solid waste before it enters the furnace, thus avoiding the difficulty in measuring the calorific value and moisture content due to the mixed composition of municipal solid waste.

[0015] Second, the active disturbance rejection controller uses the equivalent fuel disturbance as feedforward compensation, and cancels the disturbance before the furnace heat release deviates significantly, so that the furnace heat release remains near the set value when the calorific value of municipal solid waste fluctuates greatly. Compared with the feedback adjustment based on the already formed deviation, the action to suppress disturbance is more timely.

[0016] Third, the present invention sends the same equivalent fuel disturbance as a shared feedforward quantity to both the grate combustion side and the waste heat boiler side, so that the waste heat boiler side can adjust the feedwater flow rate and desuperheating water flow rate in advance before the change in furnace heat release is transmitted to the steam parameters, thus transforming the grate combustion side and the waste heat boiler side from being independent to coordinating actions with the same disturbance quantity as the link.

[0017] Fourth, both the feedwater control loop and the superheated steam temperature control loop adopt a combination of feedforward and feedback. The feedforward is responsible for the early coordination based on disturbance information, while the feedback is responsible for suppressing the lag and slow-changing disturbances on the waste heat boiler side. This ensures that the steam flow, steam pressure and superheated steam temperature remain stable when the calorific value of municipal solid waste fluctuates, thereby improving the stability of the continuous operation of the incineration line. Attached Figure Description

[0018] Figure 1 This is a schematic diagram showing the changes in the measured value of the furnace temperature and the predicted value of the expansion state observer over time in this invention; Figure 2 This is a schematic diagram illustrating the online estimation of the deviation of the equivalent fuel disturbance from the actual fuel energy in this invention. Figure 3 This is a schematic diagram illustrating the decrease in the amount of municipal solid waste fed into the furnace as the amount of equivalent fuel disturbance increases, according to the present invention. Figure 4 This is a schematic diagram illustrating how the amount of municipal solid waste fed into the furnace is adjusted according to the equivalent fuel disturbance to maintain the furnace heat release at a set value in this invention. Figure 5 This is a schematic diagram of the advance amount obtained by sharing the feedforward across the same equivalent fuel disturbance in this invention. Figure 6 This is a schematic diagram illustrating how cross-side shared feedforward suppresses superheated steam temperature fluctuations in this invention. Figure 7 This is a schematic diagram illustrating the rolling time-domain optimization principle of the upper-level model predictive control in this invention; Figure 8 This is a schematic diagram illustrating the linkage adjustment effect of steam pressure under fluctuations in the calorific value of municipal solid waste in this invention. Detailed Implementation

[0019] The following description is intended to disclose the invention and enable those skilled in the art to implement it. The preferred embodiments described below are merely examples, and other obvious variations will occur to those skilled in the art.

[0020] A method for linking the combustion of municipal solid waste grate with the load of a waste heat boiler for continuous operation, including municipal solid waste receiving, storage, feeding, grate combustion, and heat absorption and steam generation by the waste heat boiler heating surface, includes the following steps: Step 1: Real-time acquisition of operating data from the grate combustion side and the waste heat boiler side; setting up an upper-level model predictive controller, which uses the thermal inertia relationship between grate combustion and waste heat boiler as the predictive model to perform rolling prediction and rolling optimization of steam parameters, generate steam load setting trajectory, and decompose the steam load setting trajectory into furnace heat release setting value sent to the grate combustion side, feedwater setting value sent to the waste heat boiler side, and superheated steam temperature setting value sent to the waste heat boiler side. Step 2: An expansion state observer and an active disturbance rejection controller are installed on the grate combustion side. The expansion state observer takes the grate speed execution command, the primary air volume execution command, and the furnace temperature as inputs, and outputs the furnace heat release estimate and the total disturbance estimate. The expansion state observer unifies the fluctuations in the calorific value of municipal solid waste, the fluctuations in the moisture content of municipal solid waste, and other disturbances during the grate combustion process into a total disturbance acting on the grate combustion process, and uses the total disturbance estimate as an equivalent fuel disturbance. The active disturbance rejection controller forms the execution quantity on the grate combustion side based on the furnace heat release setpoint, the furnace heat release estimate, and the equivalent fuel disturbance. The execution quantity generates a grate speed execution command to adjust the amount of municipal solid waste fed into the furnace, so that when the equivalent fuel disturbance increases, the amount of municipal solid waste fed into the furnace decreases, and when the equivalent fuel disturbance decreases, the amount of municipal solid waste fed into the furnace increases. Step 3: The equivalent fuel disturbance output from the expanded state observer is used as a shared feedforward quantity and simultaneously sent to the feedwater control loop and superheated steam temperature control loop on the waste heat boiler side. The feedwater control loop adjusts the feedwater flow rate according to the feedwater setpoint and the equivalent fuel disturbance quantity, and the superheated steam temperature control loop adjusts the desuperheating water flow rate according to the superheated steam temperature setpoint and the equivalent fuel disturbance quantity. Before the change in furnace heat release is transmitted to the steam parameters, the waste heat boiler side adjusts the feedwater flow rate and desuperheating water flow rate in advance according to the equivalent fuel disturbance quantity. The grate combustion side and the waste heat boiler side use the same equivalent fuel disturbance quantity as a shared link to keep the steam parameters stable when the calorific value of municipal solid waste fluctuates, thereby realizing the linkage regulation of municipal solid waste incinerator grate combustion and waste heat boiler load.

[0021] In one specific implementation of the method for linking the grate combustion of municipal solid waste incinerator with the load of the waste heat boiler, the data for the entire continuous operation process is acquired from the distributed control system of the incineration line. Considering that grate combustion is a thermodynamic process with large inertia and long time delay, from the time municipal solid waste is pushed into the furnace to the time when the corresponding heat is converted into steam through the heating surface of the waste heat boiler, it usually takes tens of seconds or even several minutes for the transfer and accumulation to occur. Therefore, an excessively high sampling frequency is not required. One feasible approach is to collect data on transient process parameters reflecting combustion and steam production, such as furnace temperature, flue gas oxygen content, carbon monoxide concentration, steam flow rate, steam pressure, and superheated steam temperature, with a sampling period of 1 second. Feedback parameters from actuators, such as grate speed, primary air volume, and secondary air volume, should also be collected with a 1-second sampling period. For parameters reflecting the feed into the furnace, such as the weight of municipal solid waste entering the plant, the amount of waste stored in designated areas, the amount of waste grabbed by the crane, and the frequency of municipal solid waste feeding, since these parameters change slowly and occur in batches, a sampling period of 5 to 10 seconds is sufficient. This approach preserves the trend of changes on the supply side while avoiding the burden of meaningless high-frequency data collection on slowly changing signals. Regarding the ash condition of the waste heat boiler's heating surface, since ash accumulation is a slow process measured in hours, a sampling period of 60 seconds is sufficient to collect data on the flue gas and working fluid temperatures on both sides of the heating surface and characterize the degree of ash accumulation, thus adequately tracking its changes. Using different sampling periods for quantities of different natures is because when the sampling period matches the time scale of the quantity being sampled, it can ensure that no signal distortion occurs in the frequency band of interest to the control, while reserving communication and computing resources for the truly fast-changing process quantities that play a decisive role in control. This is especially important in engineering sites where dozens of signals are collected simultaneously on a single incineration line.

[0022] Before entering subsequent stages, the collected data undergoes amplitude limiting and first-order low-pass filtering. Amplitude limiting is used to eliminate abrupt values ​​that significantly exceed the physical range. For example, if the furnace temperature reading suddenly drops to near zero at a certain sampling moment, it is obviously caused by a momentary poor contact of the thermocouple. However, the thermal inertia of the furnace determines that the actual furnace temperature cannot change so drastically within 1 second. Therefore, such points are processed using the previous effective value. First-order low-pass filtering is used to suppress high-frequency measurement noise superimposed on the process quantity. Taking furnace temperature as an example, a filtering time constant of 3 seconds can be selected. This value is significantly smaller than the time constant of the furnace thermal process itself, which is on the order of tens of seconds. Therefore, while filtering out noise spikes, it does not smooth out the actual temperature change trend, thus retaining the effective information needed for subsequent stages to determine the combustion intensity.

[0023] After data acquisition and preprocessing, a predictive controller for the upper-level model is set up. The core of this controller is a predictive model describing the thermal inertia relationship between grate incineration and the waste heat boiler. Instead of establishing a complex mathematical model encompassing all mechanistic processes such as drying, pyrolysis, coke combustion, and radiative-convective heat transfer, this approach avoids the highly heterogeneous composition of municipal solid waste and the random fluctuations in calorific value and moisture content. Such a model would be difficult to calibrate accurately and maintain on-site due to the large number of parameters. A more pragmatic and robust approach to load regulation is to characterize this thermal inertia process using the dynamic response relationship between input and output: changes in furnace heat release are considered the input, and changes in three steam parameters—steam flow rate, steam pressure, and superheated steam temperature—are considered the output. The response curves between these parameters are measured through on-site step tests, and a discrete step response sequence is used as the predictive model. Specifically, when the incineration line is operating under relatively stable conditions, a small step change is made to the control parameters corresponding to the furnace heat release. The trajectories of steam flow rate, steam pressure, and superheated steam temperature over time are recorded. These three trajectories characterize the delay time and response amplitude of the furnace heat release change in the three steam parameters. For example, a step test might show that after a step increase in furnace heat release, the steam flow rate begins to rise significantly with a lag of approximately 30 seconds, reaching a new steady state after about 180 seconds. The steam pressure response is slower than the steam flow rate, and the superheated steam temperature, due to the added heat exchange components at the heating surface, typically has a longer response delay and takes even longer to reach steady state. Discretizing these measured response curves into a finite number of values ​​according to the sampling period yields all the information needed for the prediction model. This approach of using measured dynamic response relationships as a predictive model is feasible because load linkage regulation is not really concerned with the details of the temperature field at every point inside the furnace, but rather with how long and to what extent the change in furnace heat release is reflected in the steam parameters, which is precisely what the step response curve directly provides.

[0024] Within each scheduling cycle, the upper-level model predictive controller performs a rolling prediction and rolling optimization. A feasible scheduling cycle is 10 seconds. Rolling prediction refers to the controller using the measured values ​​of the three steam parameters at the current moment as a starting point, based on the aforementioned prediction model, and in conjunction with the furnace heat release changes to be taken in the subsequent steps, to gradually calculate the trends of steam flow, steam pressure, and superheated steam temperature over a future period. This predicted future time is called the prediction time domain. A feasible value is to set the prediction time domain to 180 seconds. This length is comparable to the time required for the steam flow to reach a new steady state in the aforementioned step test. Its purpose is to ensure that the prediction window is sufficient to cover the entire process of the furnace heat release changes affecting the steam parameters, so that the controller can anticipate the consequences of the current action after the delay when making decisions, rather than passively responding only after the steam parameters have deviated. This is a key arrangement for the long time lag characteristics of grate combustion, because for a process with a delay of several minutes, any immediate feedback adjustment based solely on the current deviation will cause large fluctuations in steam parameters because it will be too late.

[0025] Rolling optimization is performed based on predictions. Within each scheduling cycle, the controller uses the adjustable furnace heat release variation sequence over a future period as the optimization target. With the goal of bringing the predicted steam flow rate, steam pressure, and superheated steam temperature close to the steam load demand, an optimal furnace heat release variation sequence is solved. The solution also considers the capability limitations of the actuators, such as the adjustable range of the grate speed and its maximum variation per unit time, to ensure the result is physically feasible. The objective used in the solution can be expressed as... ; in, Let the objective value be the one to be solved. To determine the number of steps included in the prediction time domain, in this embodiment, the prediction time domain is 180 seconds, and the sampling and prediction step size is 10 seconds. Take 18; The future given by the prediction model Predicted values ​​of steam parameters; For the future The expected values ​​of the steam parameters, i.e., the targets determined by the steam load demand; That is, the future The prediction deviation of each step is squared and then accumulated over the entire prediction time domain to measure the overall degree to which the predicted steam parameter trend deviates from the expected trend. To control the number of steps included in the time domain, it means only the previous steps in the future are considered. One feasible value for adjusting the furnace heat release step by step and then keeping it constant is to set the value to [value to be filled in]. The value is set to 3, meaning that operations are only scheduled within the most recent 30 seconds. This preserves sufficient adjustment capability while avoiding increasing the solution burden and causing oscillations in the number of operations due to too many optimization variables. For the future The change in operating parameters corresponding to the heat release in the furnace step; This is a coefficient added to the change in operating parameters; its purpose is to strike a balance between quickly eliminating steam parameter deviations and avoiding overly drastic operational actions. The larger the value, the smoother the operation. The smaller the value, the more aggressive the operation. In engineering, this can be gradually adjusted during the commissioning phase, starting from small values ​​and gradually increasing, until the steam parameters can quickly keep up with the load demand and the operating mechanism does not exhibit frequent and large-scale movements. After obtaining the optimal furnace heat release change sequence, only the first step corresponding to the current scheduling cycle is executed, while the other steps are discarded. In the next scheduling cycle, the prediction is restarted, the solution is re-solved, and only the first step is executed again, starting from the newly measured steam parameters at that time. This mechanism of rolling forward one step each cycle and executing only the most recent step is called rolling optimization. Its advantage is that each decision is based on the latest measured feedback, which can continuously incorporate various unpredictable factors such as fluctuations in the calorific value of municipal solid waste and changes in external load into the correction. This gives the controller a continuous self-correcting ability to respond to disturbances, which is far more robust than calculating the entire future action at once and mechanically executing it.

[0026] See Figure 7 , Figure 7 This illustrates the rolling time-domain optimization principle of the upper-level model predictive control, as well as the relationship between the control time domain and the prediction time domain. Figure 7 It can be seen that the controller takes the current moment as the starting point, calculates the trend of steam parameters based on the prediction model in the prediction time domain, arranges operation actions only in the control time domain, and executes only the first step corresponding to the current scheduling cycle in each scheduling cycle, discarding the other steps. In the next scheduling cycle, it re-predicts, re-solves, and re-executes only the first step with the newly measured steam parameters as the starting point. Thus, each decision is based on the latest measured feedback, and can continuously incorporate unpredictable factors such as fluctuations in the calorific value of municipal solid waste into the correction.

[0027] Through the above optimization, the controller obtains not an isolated setpoint value in each scheduling cycle, but a steam load setting trajectory that unfolds over time. This trajectory shows the gradual trend that the steam parameters should follow from the current moment to a future period. A smoothly unfolding trajectory is used instead of a step-like target value because when the load demand on the grid side changes significantly, directly pushing the steam parameters to the target in one step would require a sudden and substantial change in furnace heat release. This is neither feasible for the high-inertia grate combustion process and would also cause drastic fluctuations in steam pressure. Conversely, by distributing the target change across a smooth trajectory to gradually approach it, a large disturbance is resolved into several small adjustments commensurate with the process response capability, ensuring that the steam pressure remains within the set range.

[0028] After obtaining the steam load set trajectory, the upper-level model predictive controller decomposes this trajectory into three set values, each sent to a different stage, based on the energy correspondence between steam production from the waste heat boiler and heat release from the furnace. The physical basis for this decomposition is that, from an energy perspective, the heat released by the combustion of municipal solid waste on the grate is used partly to heat the feedwater and evaporate it into saturated steam, and partly to further heat the saturated steam into superheated steam with a certain temperature; the former mainly determines the amount of steam produced, corresponding to the steam flow rate and steam pressure, while the latter mainly determines the quality of the steam, corresponding to the superheated steam temperature. Therefore, the steam production quantity requirement in the steam load setting trajectory is converted into the furnace heat release setting value that the grate combustion side should achieve, because ultimately, the amount of furnace heat release determines how much steam can be generated per unit time. The requirement in the steam load setting trajectory for maintaining the material and pressure balance of the steam-water system is converted into the feedwater setting value on the waste heat boiler side, so that the feedwater flow into the waste heat boiler matches the steam production, thereby maintaining the stability of the boiler drum water level and steam pressure. The steam quality requirement in the steam load setting trajectory is converted into the superheated steam temperature setting value on the waste heat boiler side, serving as the basis for subsequent adjustments to the desuperheating water. All three are target values ​​given by the optimization layer and are respectively sent to the grate combustion side and the waste heat boiler side. This approach of decomposing a single load trajectory into three targets—heat, feedwater, and steam temperature—based on energy destination is significant because load regulation is an energy supply and demand matching issue. Only by simultaneously clarifying and allocating the targets of the three energy sources—furnace heat release, feedwater (working fluid inlet), and superheated steam temperature (quality outlet)—can the combustion side and waste heat boiler side operate with clear objectives, preventing a situation where the combustion side only focuses on steam production quantity while the waste heat boiler side suffers from steam temperature and pressure imbalance due to insufficient feedwater or desuperheating water.

[0029] As an optional implementation, when the grid-side load demand remains constant over a certain period, the steam load setpoint trajectory can degenerate into a horizontal constant value trajectory. In this case, the aforementioned rolling prediction and rolling optimization mechanisms still operate normally, but their function changes from guiding steam parameters to follow the changing load to suppressing steam parameter drift caused by fluctuations in municipal solid waste quality under constant load. There is no difference in mechanism between the two; the only difference is whether the expected value changes over time. As another optional implementation, the specific values ​​of the aforementioned sampling period, filter time constant, scheduling period, prediction time domain, and control time domain can all be adjusted within the corresponding range according to the actual inertia of the incineration line and the characteristics of load fluctuations. For example, for incineration lines with greater thermal inertia and higher single-unit processing capacity, the prediction time domain can be made longer to ensure that the prediction window can always cover the complete process of furnace heat release changes being transmitted to steam parameters. All such adjustments do not change the basic implementation methods of the aforementioned rolling prediction, rolling optimization, and decomposition of setpoints according to energy correspondence.

[0030] In one specific implementation of the method for linking the grate combustion and waste heat boiler load regulation in municipal solid waste incinerators, the regulation on the grate combustion side is jointly accomplished by an extended state observer and an active disturbance rejection controller. To understand the implementation of this part, it is essential to first recognize the fundamental difficulty faced by grate incineration: the amount of heat released per unit time by municipal solid waste on the grate depends on the calorific value and moisture content of the waste entering the furnace. However, municipal solid waste is highly mixed in composition, its calorific value fluctuates randomly within a wide range of 4000 kJ / kg to 9000 kJ / kg, and its moisture content is unevenly distributed between 40% and 60%. These two quantities cannot be directly measured before entering the furnace and are constantly changing. Therefore, this implementation does not attempt to measure the calorific value and moisture content online. Instead, it adopts a different approach, treating their entire impact on the combustion process as a disturbance that needs to be estimated. This is the starting point for setting up the extended state observer.

[0031] The expansion state observer selects furnace temperature as an observable measure reflecting furnace heat release, which is the premise for the entire estimation to be valid. Furnace temperature can be used to reflect furnace heat release because, under relatively stable furnace structure and flue gas flow conditions, the more heat released by the combustion of municipal solid waste on the grate, the more fully the flue gas in the furnace is heated, and the higher the furnace temperature will be. There is a stable and clear correspondence between the two. The furnace temperature is continuously measured by thermocouples placed in appropriate positions in the furnace with the aforementioned 1-second sampling period. It is a quantity that is always available on site, so using it as an observable measure of furnace heat release is both realistic and reliable. Based on this correspondence, the grate combustion process can be approximated as a second-order dynamic process: the furnace heat release itself is the first state, the rate of change of furnace heat release over time is the second state, and the grate speed execution command and the primary air volume execution command are the inputs that drive this process. The grate speed execution command changes the amount of domestic waste pushed into the furnace per unit time to change the amount of fuel available for combustion, and the primary air volume execution command changes the intensity of combustion by changing the amount of oxygen supplied for combustion.

[0032] See Figure 1 , Figure 1 The relationship between the measured values ​​of furnace temperature and the predicted values ​​from the expansion state observer is shown over time. Figure 1 It can be seen that during the process of the calorific value of municipal solid waste increasing and decreasing, the furnace temperature predicted by the observer always closely matches the measured furnace temperature. This indicates that the observer can correct the internal furnace heat release estimate, furnace heat release rate estimate, and total disturbance estimate based on the observation deviation, so that the observation deviation converges as the control cycle progresses.

[0033] The key to the extended state observer lies in its addition of a third state beyond the two states mentioned above, specifically designed to accommodate all factors not described by the nominal model. Specifically, the nominal model only characterizes the impact of grate speed and primary air volume commands on furnace temperature under a given quality of municipal solid waste—that is, the nominal input-output relationship of the grate combustion process. The portion of the actual furnace heat release that differs from the nominal model's prediction, caused by factors such as deviations in calorific value and moisture content of municipal solid waste, and incomplete combustion, is all categorized into this third extended state. It is inherently reasonable to group factors from different sources, such as calorific value fluctuations, moisture content fluctuations, and other disturbances during grate combustion, into a single expansion state. Although these factors have different causes, they affect furnace heat release through the same pathway: the burning municipal solid waste on the grate, ultimately manifesting as a deviation in furnace heat release from the nominal value. Since their pathway and result are consistent, they can be uniformly characterized by a single aggregate quantity, without needing to, and indeed without being able to, separate them individually. The estimated value of this aggregate expansion state is taken as the equivalent fuel disturbance quantity.

[0034] The operation of the extended state observer can be represented as ; in, For the first The state estimate maintained within the observer of each control cycle consists of three components: the furnace heat release estimate, the furnace heat release rate estimate, and the total disturbance estimate. State estimation for the next control cycle; To control the cycle, it is set to 1 second in this embodiment; The coefficient matrix describes the internal evolution relationship of the aforementioned second-order process. It characterizes the recursive relationship between furnace heat release, its rate of change, and expansion state over time. For the first Each control cycle acts on the input of the grate combustion process, consisting of grate speed execution commands and primary air volume execution commands; The coefficient matrix is ​​the input that acts on the process. It is determined by the nominal input-output correspondence of the aforementioned grate combustion process and describes the degree of influence of the grate speed execution command and the primary air volume execution command on the furnace heat release. The actual furnace temperature collected during the first control cycle; This is the predicted value of the furnace temperature estimated and calculated by the observer based on its internal state; The observation bias represents the difference between the measured furnace temperature and the furnace temperature predicted by the observer. The correction gain of the observer determines the strength of the feedback correction of the three state estimates by the observation bias. This relationship means that each time the observer completes a control cycle, it first extrapolates one step forward based on its current state estimate and known inputs, according to the internal evolutionary relationship of the process. Then, it corrects the extrapolation result using the observation bias between the measured furnace temperature and the predicted furnace temperature, and so on, repeating this process. The correction step is crucial because it is precisely by continuously using the observation bias to influence the three estimates, including the expansion state, that the observer can estimate the equivalent fuel disturbance, which cannot be directly measured, even without knowing the specific values ​​of calorific value and moisture content, relying solely on the three known quantities: grate speed execution command, primary air volume execution command, and furnace temperature.

[0035] Correction gain The value of is directly related to the speed and stability of the estimation. The larger the value, the stronger the feedback of the observation bias, and the faster the observer tracks the real state. However, it is also easier to introduce the measurement noise superimposed on the furnace temperature as a real change into the estimate, causing the estimated value of the equivalent fuel disturbance to fluctuate at high frequency. Conversely, a smaller bandwidth results in a more stable estimate but slower tracking, meaning the observer takes longer to reflect sudden changes in the calorific value of municipal solid waste. An engineering approach to facilitate tuning is to unify the correction strength for the three observer states using a single observer bandwidth. A higher bandwidth leads to stronger overall correction and faster tracking, while a lower bandwidth results in greater overall stability. In this implementation, the observer bandwidth can be initially tuned so that the equivalent fuel disturbance can be reflected within approximately 10 seconds after a step change in the calorific value of municipal solid waste. Then, it can be adjusted appropriately based on the stability of the estimated equivalent fuel disturbance until both tracking speed and stability are achieved. The output equivalent fuel disturbance numerically represents the deviation trend of the actual energy released by the municipal solid waste per unit time relative to the energy required to maintain the furnace heat release setpoint: a positive and continuously increasing value indicates that the actual energy released by the current municipal solid waste is higher than the energy required to maintain the setpoint, while a negative value indicates that the actual released energy is lower than the required energy.

[0036] See Figure 2 , Figure 2 This illustrates the relationship between the equivalent fuel disturbance and the actual fuel energy deviation for online estimation. Figure 2 It can be seen that after the calorific value of municipal solid waste undergoes a step change, the equivalent fuel disturbance can track the actual fuel energy deviation in a short period of time. Thus, even without knowing the specific values ​​of calorific value and moisture content, the equivalent fuel disturbance, which cannot be directly measured, can be estimated based solely on the grate speed execution command, the primary air volume execution command, and the furnace temperature.

[0037] After obtaining the estimated furnace heat release and equivalent fuel disturbance, the Active Disturbance Rejection Controller (ADRC) formulates the execution parameters for the grate combustion side. The ADRC first uses a tracking differentiator to arrange a transition process for the furnace heat release setpoint. The furnace heat release is not allowed to directly chase the step change in the setpoint because the setpoint sent from the upper level may experience large jumps during load changes. Grate combustion is a process with high inertia; rigidly chasing a step target would only induce violent grate speed movements and cause overshoot in the furnace heat release. The tracking differentiator transforms the step change in the setpoint into a transitional furnace heat release setpoint that smoothly transitions from the current value to the target value, while simultaneously providing the trend of this transitional setpoint over time. This ensures that subsequent adjustments have a target that is commensurate with the process response capability and whose rate of change is known.

[0038] Subsequently, the active disturbance rejection controller (ADRC) uses the deviation between the furnace heat release transition setpoint and the estimated furnace heat release as the tracking deviation. Based on this tracking deviation and its changing trend, it generates a feedback adjustment quantity. The larger the tracking deviation, the larger the feedback adjustment quantity; the faster the tracking deviation changes, the larger the feedback adjustment quantity. This part performs the function of conventional closed-loop adjustment based on the current tracking situation. What truly distinguishes this method from ordinary closed-loop adjustment is the subsequent feedforward compensation stage: the ADRC converts the equivalent fuel disturbance into a feedforward compensation quantity to pre-counteract the effect of the equivalent fuel disturbance on furnace heat release. Then, it superimposes the feedback adjustment quantity and the feedforward compensation quantity to obtain the final execution quantity on the grate combustion side. The composition relationship can be expressed as follows: ; in, For the first The execution quantity of the grate combustion side in each control cycle; The furnace heat release transition setpoint for this cycle; This is the estimated value of the furnace heat release for this cycle, which is the first of the three estimates of the aforementioned observer; That is, tracking deviation; The trend of the furnace heat release transition setpoint changing over time; The feedback coefficient that acts on the tracking deviation. The larger the gain, the stronger the adjustment for the same tracking deviation; As a feedback coefficient acting on the trend of change in the transition setpoint, it enables the execution quantity to take into account the speed of change of the target and respond in advance rather than always lagging behind the target; This is the equivalent fuel disturbance for that period, i.e., the expansion state estimated by the aforementioned observer; The nominal value representing the degree of influence of the grate combustion-side input on the furnace heat release is derived from the aforementioned nominal input-output correspondence. This involves converting the equivalent fuel disturbance, characterized by deviations in furnace heat release, into the required reverse adjustment in the execution quantity to offset it. The minus sign before this term is crucial for feedforward compensation. The sum of the first two terms constitutes the feedback adjustment, and the third term is the feedforward compensation. The sum of these two terms forms the execution quantity, which is entirely consistent with the setting of the active disturbance rejection controller in this method.

[0039] Here's an arrangement that seems counterintuitive but precisely reflects the method's characteristics: why the execution quantity changes towards reducing the amount of municipal solid waste fed into the furnace when the equivalent fuel disturbance increases. Imagine a scenario where the calorific value of municipal solid waste suddenly increases: the same amount of waste fed into the furnace will now release more heat than under standard conditions, causing the furnace heat release to rise and leading to an increase in furnace temperature. The equivalent fuel disturbance estimated by the expansion state observer will thus increase. If, in the naive approach of increasing the amount of waste fed into the furnace based on the usual consideration of steam production, this would only further increase the already high furnace heat release, leading to furnace overheating and a significant surge in steam parameters. The correct approach is precisely the opposite: since a unit mass of municipal solid waste can release more heat at this moment, to maintain the furnace heat release at the set value, the amount of waste fed into the furnace should be reduced. With less material and its increased calorific value, the amount of heat released should be exactly equal to the set value. The minus sign preceding the equivalent fuel disturbance term in the aforementioned synthetic relationship solidifies this physical fact into the generation of the execution quantity: as the equivalent fuel disturbance increases, the minus sign causes the execution quantity to decrease, thereby reducing the amount of municipal solid waste fed into the furnace through subsequent stages; as the equivalent fuel disturbance decreases, meaning the calorific value of the municipal solid waste is low and the same amount of material is insufficient for heat release, the execution quantity increases to make up for the heat gap with more material. Thus, the furnace heat release is maintained near the set value even with significant fluctuations in the calorific value of the municipal solid waste. Feedforward compensation is superior to simple feedback because the expansion state observer estimates the equivalent fuel disturbance before the furnace heat release deviates significantly, when only the furnace temperature shows initial signs of change. Therefore, the execution quantity can preemptively compensate before the deviation fully develops, suppressing the disturbance at its nascent stage. Simple feedback regulation, on the other hand, can only act after the deviation has actually formed and been detected. For the already slow grate combustion process, this difference in timing often marks the difference between stable and drastic fluctuations in steam parameters.

[0040] See Figure 3 and Figure 4 , Figure 3 The diagram shows the relationship between the decrease in the amount of municipal solid waste fed into the furnace as the equivalent fuel disturbance increases. Figure 4 This illustrates the relationship between adjusting the amount of municipal solid waste fed into the furnace according to the equivalent fuel disturbance to maintain the furnace heat release at a set value. Figure 3It can be seen that when the calorific value of municipal solid waste increases, leading to a greater disturbance in the equivalent fuel, the amount of municipal solid waste fed into the furnace decreases accordingly; from Figure 4 It can be seen that when the feed is reduced according to the equivalent fuel disturbance, the furnace heat release is maintained near the set value. However, when the feed amount is maintained without compensation, the furnace heat release surges significantly. The difference between the two is the overshoot avoided by reducing the feed. Reducing the amount of municipal solid waste fed into the furnace when the equivalent fuel disturbance increases is because the calorific value of municipal solid waste increases, and the same amount of municipal solid waste fed into the furnace will release more heat. In order to maintain the furnace heat release at the set value, less municipal solid waste should be fed into the furnace, and its increased calorific value should release exactly the heat required to reach the set value.

[0041] After the anti-disturbance controller obtains the execution quantity of the grate combustion side based on the above-mentioned synthesis relationship, it generates a grate speed execution command from this execution quantity. The grate speed execution command drives the grate hydraulic drive mechanism to change the grate speed, thereby changing the amount of domestic waste pushed into the furnace for combustion per unit time, thus realizing the aforementioned adjustment intention. At the same time, in order to ensure that the changed amount of domestic waste can still be fully burned, the primary air volume execution command changes in the same direction as the grate speed execution command. That is, when the amount of domestic waste entering the furnace increases, the primary air volume increases accordingly, and when the amount of domestic waste entering the furnace decreases, the primary air volume decreases accordingly, so as to maintain the ratio between the primary air volume and the amount of domestic waste entering the furnace. The primary air volume is also corrected according to the oxygen content of the flue gas to keep the oxygen content of the flue gas within a suitable range for combustion. The secondary air volume execution command is formed based on the oxygen content of the flue gas and the carbon monoxide concentration. When the carbon monoxide concentration increases, indicating that the combustion is not complete, the secondary air volume is increased to enhance combustion and thus maintain the degree of combustion of domestic waste on the grate.

[0042] As an optional implementation, the aforementioned treatment of treating the grate combustion process as a second-order dynamic process can be adjusted to a first-order dynamic process based on the actual response characteristics of the grate combustion on site. That is, only the furnace heat release state is considered, and its rate of change is no longer set separately. Correspondingly, the expansion state observer is reduced from maintaining three estimates to maintaining two estimates. The remaining implementation methods—correcting the observation deviation state, incorporating disturbances into the expansion state and taking them as equivalent fuel disturbances, and then forming the execution quantities—remain unchanged. This simplification is easier to tune in situations where the grate combustion response is relatively smooth, at the cost of slightly reducing the ability to track rapid changes. A trade-off can be made depending on the specific combustion line. As another optional implementation, the aforementioned observer bandwidth and feedback coefficient... and The specific values ​​can be gradually adjusted during the on-site commissioning phase based on the speed and stability of the furnace heat release transition setpoint. All such adjustments do not change the basic implementation method of estimating the equivalent fuel disturbance amount by the expansion state observer and synthesizing the grate combustion side execution amount by the active disturbance rejection controller based on the furnace heat release setpoint, the furnace heat release estimate, and the equivalent fuel disturbance amount.

[0043] In one specific implementation of the method for linking the grate combustion and waste heat boiler load regulation in municipal solid waste incinerators, the load regulation on the waste heat boiler side is based on the equivalent fuel disturbance estimated by the expansion state observer on the grate combustion side. The feedwater control and superheated steam temperature control on the waste heat boiler side typically form closed loops based on measured steam parameters. That is, the feedwater flow rate or desuperheating water flow rate is adjusted only after the steam pressure or superheated steam temperature has deviated from the target. However, the disturbance caused by fluctuations in the calorific value of municipal solid waste originates far in the combustion stage on the grate. The excess or deficiency of heat released must first undergo continuous combustion of the municipal solid waste on the grate, then be transferred through various stages of heating surfaces with the flue gas flow, and finally be reflected in changes in steam flow rate, steam pressure, and superheated steam temperature. This transmission process often takes tens of seconds or even minutes. By the time the waste heat boiler side detects an anomaly in the steam parameters, the heat causing the anomaly has already traveled its transmission route or even partially arrived. Adjusting the feedwater and desuperheating water at this point will inevitably be delayed. To fundamentally resolve this lag, this implementation method uses the equivalent fuel disturbance output from the grate combustion side expansion state observer as a shared feedforward quantity, which is simultaneously sent to the feedwater control loop and the superheated steam temperature control loop on the waste heat boiler side.

[0044] See Figure 5 , Figure 5 This illustrates the lead time obtained by sharing feedforward across the same equivalent fuel disturbance, as well as the temporal sequence of the equivalent fuel disturbance, the arrival of steam heat at the heating surface, and the feedforward actions of feedwater and desuperheating water. Figure 5 It can be seen that the equivalent fuel disturbance is estimated by the expansion state observer as soon as the furnace temperature begins to change and before the steam parameters are affected. This estimated heat deviation only becomes apparent in the steam parameters after combustion and heat exchange. There is a time interval between the equivalent fuel disturbance and the initial rise of steam heat reaching the heating surface; this interval is the advance gained by the waste heat boiler. This advance is not a prediction of the future, but rather utilizes the objective time interval between the disturbance being observed on the grate combustion side and its consequences not yet manifesting on the waste heat boiler side. Using this advance, the waste heat boiler can adjust its heat absorption capacity before the excess or deficiency of steam heat actually reaches the heating surface, thus ensuring a smooth transition of steam parameters when the heat subsequently arrives.

[0045] The fact that the same equivalent fuel disturbance can simultaneously drive two different loops—the feedwater control loop and the superheated steam temperature control loop—is determined by the objects they govern and the mode of action of the disturbance. The feedwater control loop mainly concerns the steam production quantity and the material and pressure balance of the steam-water system, corresponding to steam flow rate and steam pressure; the superheated steam temperature control loop mainly concerns steam quality, corresponding to superheated steam temperature. Fluctuations in the calorific value of municipal solid waste causing deviations in furnace heat release simultaneously impact both steam production quantity and steam quality. Therefore, using the same equivalent fuel disturbance as the common feedforward source for both loops is logically sound. Considering that changes in feedwater flow rate and desuperheating water flow rate affect steam parameters at different rates—superheated steam temperature typically responds more rapidly to desuperheating water flow rate, while steam pressure responds relatively slowly to feedwater flow rate—this implementation allows each loop to convert the same equivalent fuel disturbance into feedwater feedforward correction and desuperheating water feedforward correction quantities, respectively, based on their own response characteristics. The feedwater control loop adds a feedwater feedforward correction to the feedwater feedback regulation and adjusts the feedwater flow accordingly. The superheated steam temperature control loop adds a desuperheating water feedforward correction to the desuperheating water feedback regulation and adjusts the desuperheating water flow accordingly. The two feedforward corrections are formed independently from equivalent fuel disturbances of the same value. They share the same disturbance information and avoid over-adjustment of superheated steam temperature or oscillation on the feedwater side caused by forcibly making the two corrections equal.

[0046] The feedforward correction and feedback regulation work in two separate loops, demonstrating that the waste heat boiler side does not rely solely on the equivalent fuel disturbance. The feedforward correction is responsible for acting proactively based on disturbance information from the grate combustion side, compensating for deviations caused by external sources such as fluctuations in the calorific value of municipal solid waste before they materialize. Feedwater feedback regulation and desuperheating water feedback regulation, on the other hand, are responsible for suppressing disturbances inherent to the waste heat boiler side itself, such as the inherent water-side lag in the feedwater pipeline and steam-water system, changes in steam demand at the steam end, and changes in heat exchange capacity caused by changes in the ash state of the waste heat boiler's heating surface. The ash accumulation on the heating surface is a slow process measured in hours, which is tracked and eliminated by the desuperheating water feedback regulation based on the measured superheated steam temperature. In this way, the disturbances from the combustion source in the feedforward tube and the lag and gradual changes in the waste heat boiler side in the feedback tube each perform their respective functions and complement each other, ensuring that the regulation on the waste heat boiler side is both timely and stable.

[0047] See Figure 6 , Figure 6 The relationship between cross-side shared feedforward and the suppression of superheated steam temperature fluctuations is shown. Figure 6It can be seen that without cross-side shared feedforward, the waste heat boiler side only adjusts based on feedback after the superheated steam temperature deviates, resulting in a significant upward surge in superheated steam temperature. However, with cross-side shared feedforward, the desuperheating water is pre-set based on the equivalent fuel disturbance before the extra steam heat reaches the heating surface, significantly reducing the deviation of superheated steam temperature. The difference between the two is the overshoot reduced by introducing cross-side shared feedforward.

[0048] Thus, the grate combustion side and the waste heat boiler side work in concert around the same equivalent fuel disturbance. When the calorific value of municipal solid waste increases, the equivalent fuel disturbance increases. On the one hand, the grate combustion side reduces the amount of municipal solid waste fed into the furnace, attempting to bring the furnace heat release, which tends to rise due to the increased calorific value, back to the furnace heat release setpoint. On the other hand, during the transition period before the furnace heat release returns to its normal level, the grate will still release some extra heat temporarily, which will be transferred to the heating surface with the flue gas. At this time, the waste heat boiler side increases the feedwater flow rate and desuperheating water flow rate simultaneously based on the increased equivalent fuel disturbance. The increased feedwater is used to absorb the extra steam heat generated during this transition period and stabilize the steam pressure, while the increased desuperheating water suppresses the superheated steam temperature, which tends to rise due to the excess heat. Conversely, when the calorific value of municipal solid waste decreases, the equivalent fuel disturbance decreases. The grate combustion side adjusts the amount of municipal solid waste fed into the furnace to increase, thereby raising the decreasing furnace heat release back to the set value. Meanwhile, the waste heat boiler side simultaneously reduces the feedwater and desuperheating water flow rates based on the reduced equivalent fuel disturbance. It is evident that the feed reduction and increase on the grate combustion side, and the increase and decrease in heat absorption on the waste heat boiler side, occur simultaneously and in mutually responsive directions, based on the same equivalent fuel disturbance. The two systems work closely together, bound by this shared quantity.

[0049] See Figure 8 , Figure 8 This demonstrates the linkage effect of steam pressure regulation under fluctuations in the calorific value of municipal solid waste, as well as the set range of steam pressure. (From...) Figure 8 It can be seen that during the continuous fluctuation of the calorific value of municipal solid waste, the steam pressure fluctuates greatly and exceeds the set range many times when there is no linkage regulation. However, when the linkage regulation of the present invention is adopted, the reduction of fuel on the grate combustion side and the advance adjustment on the waste heat boiler side are carried out in coordination with the same equivalent fuel disturbance amount as the link, and the steam pressure is stably maintained within the set range, thereby realizing the linkage regulation of municipal solid waste incinerator grate combustion and waste heat boiler load.

[0050] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the claimed invention. The scope of protection claimed by the appended claims and their equivalents is defined.

Claims

1. A method for linkage regulation of grate combustion and waste heat boiler load in a municipal solid waste incinerator, applied to a continuous operation process including municipal solid waste receiving, storage, feeding, grate combustion, and waste heat boiler heat exchange surface heat absorption and steam generation, characterized in that, Includes the following steps: Step 1: Real-time acquisition of operating data from the grate combustion side and the waste heat boiler side; setting up an upper-level model predictive controller, which uses the thermal inertia relationship between grate combustion and waste heat boiler as the predictive model to perform rolling prediction and rolling optimization of steam parameters, generate steam load setting trajectory, and decompose the steam load setting trajectory into furnace heat release setting value sent to the grate combustion side, feedwater setting value sent to the waste heat boiler side, and superheated steam temperature setting value sent to the waste heat boiler side. Step 2: Set up an expansion state observer and an active disturbance rejection controller on the grate combustion side. The expansion state observer takes the grate speed execution command, the primary air volume execution command and the furnace temperature as inputs and outputs the furnace heat release estimate and the total disturbance estimate. The extended state observer integrates the fluctuations in calorific value and moisture content of municipal solid waste, as well as other disturbances during grate combustion, into a total disturbance acting on the grate combustion process, and uses the estimated total disturbance as the equivalent fuel disturbance. The active disturbance rejection controller generates the execution quantity on the grate combustion side based on the furnace heat release setpoint, the estimated furnace heat release, and the equivalent fuel disturbance. The execution quantity generates a grate speed execution command to adjust the amount of municipal solid waste fed into the furnace, so that when the equivalent fuel disturbance increases, the amount of municipal solid waste fed into the furnace decreases, and when the equivalent fuel disturbance decreases, the amount of municipal solid waste fed into the furnace increases. Step 3: The equivalent fuel disturbance output from the expanded state observer is used as a shared feedforward quantity and simultaneously sent to the feedwater control loop and superheated steam temperature control loop on the waste heat boiler side. The feedwater control loop adjusts the feedwater flow rate according to the feedwater setpoint and the equivalent fuel disturbance quantity, and the superheated steam temperature control loop adjusts the desuperheating water flow rate according to the superheated steam temperature setpoint and the equivalent fuel disturbance quantity. Before the change in furnace heat release is transmitted to the steam parameters, the waste heat boiler side adjusts the feedwater flow rate and desuperheating water flow rate in advance according to the equivalent fuel disturbance quantity. The grate combustion side and the waste heat boiler side use the same equivalent fuel disturbance quantity as a shared link to keep the steam parameters stable when the calorific value of municipal solid waste fluctuates, thereby realizing the linkage regulation of municipal solid waste incinerator grate combustion and waste heat boiler load.

2. The method according to claim 1, characterized in that, The real-time operational data collected in step 1 includes the weighing of municipal solid waste entering the plant, the storage capacity of the waste bin zones, the amount of waste grabbed by the garbage crane, the frequency of municipal solid waste feeding, the grate speed, the primary air volume, the secondary air volume, the furnace temperature, the oxygen content of the flue gas, the carbon monoxide concentration, the steam flow rate, the steam pressure, the superheated steam temperature, and the ash status of the waste heat boiler heating surface. The weighing of municipal solid waste entering the plant, the storage capacity of the waste bin zones, and the amount of waste grabbed by the garbage crane represent the supply of municipal solid waste into the furnace. The furnace temperature, the oxygen content of the flue gas, and the carbon monoxide concentration represent the combustion of municipal solid waste on the grate. The steam flow rate, the steam pressure, and the superheated steam temperature represent the steam production of the waste heat boiler. The amount of municipal solid waste entering the furnace is jointly determined by the grate speed and the frequency of municipal solid waste feeding. The grate speed execution command changes the amount of municipal solid waste pushed into the furnace for incineration per unit time by adjusting the grate speed.

3. The method according to claim 2, characterized in that, The steam parameters mentioned in step 1 include steam flow rate, steam pressure, and superheated steam temperature. The thermal inertia relationship characterizes the delay and correspondence between changes in furnace heat release and the thermal inertia of the waste heat boiler, which are transmitted to steam flow rate, steam pressure, and superheated steam temperature. In each scheduling cycle, the upper-level model predictive controller performs rolling predictions of steam flow rate, steam pressure, and superheated steam temperature for subsequent set periods based on the prediction model. It also performs rolling optimization with the goal of making the steam flow rate, steam pressure, and superheated steam temperature obtained from the rolling predictions approach the steam load demand, generating a steam load setting trajectory to maintain the steam pressure within the set range. Based on the energy correspondence between steam production from the waste heat boiler and furnace heat release, the upper-level model predictive controller decomposes the steam load setting trajectory into furnace heat release setpoints, feedwater setpoints, and superheated steam temperature setpoints. The furnace heat release setpoints, feedwater setpoints, and superheated steam temperature setpoints are all target values ​​given by the optimization layer.

4. The method according to claim 1, characterized in that, In step 2, the expansion state observer maintains three estimates internally: the furnace heat release estimate, the furnace heat release rate estimate, and the total disturbance estimate. All three estimates are observed quantities. The nominal input-output correspondence of the grate combustion process characterizes the degree of influence of grate velocity and primary air volume on furnace temperature under nominal municipal solid waste quality. In each control cycle, the extended state observer predicts the furnace temperature based on the current grate speed execution command, primary air volume execution command, and the nominal input-output correspondence of the grate combustion process. The predicted furnace temperature is compared with the furnace temperature collected in the next control cycle to obtain the observation deviation. Based on the observation deviation, the furnace heat release estimate, the furnace heat release rate estimate, and the total disturbance estimate are simultaneously corrected so that the observation deviation converges as the control cycle progresses, and the furnace heat release estimate and the total disturbance estimate approach the actual values ​​as the control cycle progresses.

5. The method according to claim 4, characterized in that, In step 2, the calorific value and moisture content that change with the quality of municipal solid waste are estimated online by the expansion state observer as quantities to be estimated; the equivalent fuel disturbance represents the deviation trend of the actual energy released by the municipal solid waste entering the furnace per unit time relative to the energy required to maintain the furnace heat release set value.

6. The method according to claim 1, characterized in that, In step 2, the active disturbance rejection controller (ADRC) is equipped with a tracking differentiator. The tracking differentiator arranges a transition process for the furnace heat release setpoint, outputs the furnace heat release transition setpoint and its changing trend, so that the step change of the furnace heat release setpoint is transformed into a smooth transition setpoint. The ADRC uses the deviation between the furnace heat release transition setpoint and the furnace heat release estimate as the tracking deviation. The tracking deviation and its changing trend form a feedback adjustment quantity. The feedback adjustment quantity increases with the increase of the tracking deviation and the increase of the rate of change of the tracking deviation. The ADRC converts the equivalent fuel disturbance quantity into a feedforward compensation quantity, which is used to offset the effect of the equivalent fuel disturbance quantity on the furnace heat release. The ADRC superimposes the feedback adjustment quantity and the feedforward compensation quantity to obtain the execution quantity on the grate combustion side.

7. The method according to claim 6, characterized in that, In step 2, the grate speed execution command drives the grate hydraulic drive mechanism to adjust the grate speed to adjust the amount of domestic waste entering the furnace; the primary air volume execution command changes in the same direction as the grate speed execution command and is corrected according to the oxygen content of the flue gas. The primary air volume is adjusted by the primary air regulating damper to maintain the ratio of primary air volume to the amount of domestic waste entering the furnace; the secondary air volume execution command is formed based on the oxygen content and carbon monoxide concentration of the flue gas. The secondary air volume is adjusted by the secondary air regulating damper to ensure the complete combustion of domestic waste on the grate.

8. The method according to claim 1, characterized in that, In step 3, both the feedwater control loop and the superheated steam temperature control loop are control loops that combine feedforward and feedback. The feedwater control loop uses the collected steam flow rate and steam pressure as feedback quantities, and forms feedwater feedback regulation based on the deviation between the feedwater setpoint and the feedback quantity. The feedwater feedback regulation suppresses the deviation caused by the lag of the waste heat boiler side water and the change in steam demand. The superheated steam temperature control loop uses the collected superheated steam temperature as feedback quantity, and forms desuperheating water feedback regulation based on the deviation between the superheated steam temperature setpoint and the feedback quantity. The desuperheating water feedback regulation suppresses the superheated steam temperature deviation caused by the change in the ash state of the heat exchange surface of the waste heat boiler.

9. The method according to claim 8, characterized in that, In step 3, the equivalent fuel disturbance is simultaneously sent to both the feedwater control loop and the superheated steam temperature control loop via the signal channel in each control cycle. This ensures that both loops receive the same equivalent fuel disturbance value within the same control cycle. Based on its own response characteristics to changes in feedwater flow, the feedwater control loop converts the received equivalent fuel disturbance into a feedwater feedforward correction. This correction increases with increasing equivalent fuel disturbance and decreases with decreasing equivalent fuel disturbance. The feedwater control loop then superimposes the feedwater feedforward correction onto the feedwater feedback regulation and adjusts the feedwater flow based on the superimposed result. Flow rate; The superheated steam temperature control loop, based on its own response characteristics to changes in desuperheating water flow rate, converts the received equivalent fuel disturbance into a desuperheating water feedforward correction. The desuperheating water feedforward correction increases with the increase of the equivalent fuel disturbance and decreases with the decrease of the equivalent fuel disturbance. The superheated steam temperature control loop superimposes the desuperheating water feedforward correction on the basis of desuperheating water feedback regulation and adjusts the desuperheating water flow rate according to the superimposed result. The feedwater feedforward correction and the desuperheating water feedforward correction are formed independently from equivalent fuel disturbances of the same value, so that the feedwater flow rate and the desuperheating water flow rate are corrected in advance according to the response characteristics of their respective loops.

10. The method according to claim 1, characterized in that, In step 3, before the changes in furnace heat release are transferred to the steam flow, steam pressure, and superheated steam temperature through the combustion of domestic waste on the grate and the thermal inertia of the waste heat boiler, the waste heat boiler side adjusts the feedwater flow and desuperheating water flow in advance based on the equivalent fuel disturbance, so that the heat absorption capacity of the waste heat boiler side is preset before the excess or deficiency of steam heat reaches the heating surface. When the furnace heat release increases, the grate combustion side changes the amount of domestic waste fed into the furnace in a decreasing direction to bring the furnace heat release back to the furnace heat release set value, and the waste heat boiler side simultaneously increases the feedwater flow and desuperheating water flow to absorb the excess steam heat generated during the furnace heat release increase phase. When the furnace heat release decreases, the grate combustion side changes the amount of domestic waste fed into the furnace in a increasing direction to bring the furnace heat release back to the furnace heat release set value, and the waste heat boiler side simultaneously decreases the feedwater flow and desuperheating water flow.

Images