Closed-loop intelligent optimization control system for boiler combustion process

Abstract

A closed-loop intelligent optimization control system for a boiler combustion process comprises four modules: a system calculation module, a system monitoring module, a system communication module and a system interface module. A system calculation module calculates the optimization quantity of secondary air, burnout air, coal feeding amount offset and oxygen amount fixed value of each layer of the hearth, and meanwhile, if the modeling error is large, an online support vector machine model is updated; in each control period, a system calculation module needs to judge whether a current optimization system is input or not, if yes, the output control quantity is calculated according to the process, and if not, the output control quantity of an original control system is tracked; boiler combustion optimization is researched by means of a neural network and the like; and an online least square support vector machine is developed to establish a boiler dynamic model system, and intelligent closed-loop dynamic combustion optimization control of the boiler is promoted.

Description

Technical Field

[0001] This invention relates to the field of industrial automatic control technology, specifically to a closed-loop intelligent optimization control system for boiler combustion process. Background Technology

[0002] With the continuous commissioning of 600 MW-class boilers, the constant changes in coal market prices and coal combustion structure, and the increasing pressure from environmental policies, improving boiler combustion efficiency and reducing NOx emissions have become paramount. Currently, the conventional DCS control system for supercritical units within the group primarily controls the total coal feed and air volume required by the boiler based on different load demands. However, boiler burners have multiple layers, and the distribution of coal and air volume among these layers affects the boiler's combustion conditions, thus impacting combustion efficiency and NOx emissions in the flue gas. Currently, the coal feed to each burner layer is generally evenly distributed. For air supply, the total air volume under different loads is typically calculated. However, boiler air supply involves fuel air, auxiliary air, burnout air, and additional air added to the furnace at different layers (heights) for combustion. Given a fixed total boiler air volume, how these different layers (heights) of air volume are distributed significantly affects the boiler's combustion conditions, thereby impacting combustion efficiency and NOx emissions in the flue gas. Currently, the allocation of air volume in conventional optimization systems is based on certain empirical rules, which has not reached the optimal level. There is room for further optimization in both the allocation of coal feed and the allocation of air volume in each layer. Summary of the Invention

[0003] In view of the above situation and to overcome the defects of the prior art, the technical solution adopted by the present invention is as follows: A closed-loop intelligent optimization control system for boiler combustion process includes four modules: a system calculation module, a system monitoring module, a system communication module, and a system interface module. Its key feature is that the system calculation module calculates the optimization quantities for secondary air, combustion air, coal feed rate bias, and oxygen setpoint at each furnace layer. Simultaneously, if the modeling error is large, the online support vector machine model is updated. In each control cycle, the system calculation module needs to determine whether the current optimization system is in operation. If so, it calculates the output control quantity according to the process; otherwise, it tracks the output control quantity of the original control system.

[0004] Compared with the prior art, the present invention has the following beneficial effects: We will study boiler combustion optimization using neural networks and other methods; develop an online least squares support vector machine to establish a boiler dynamic model system; conduct research on boiler economy and boiler efficiency; and promote intelligent closed-loop dynamic combustion optimization control of boilers.

[0005] 1. After the optimized control system is put into operation, the boiler thermal efficiency will increase by approximately 0.5%.

[0006] 2. A 600MW coal-fired unit operates for approximately 8 months of the year, totaling about 5760 hours annually (24 × 30 × 8 = 5760 hours). Assuming a normal operating coal consumption of 270g / kWh, the annual standard coal consumption before the combustion optimization system was put into operation was 5760 hours × 540MW (average annual load) × 270g / (kW·h) = 5760h × 540 × 10³ kW × 270 g / (kW·h) = 839808 tons. Therefore, the standard coal consumption would be 839808 × 93% / 93.5% = 835317 tons. If the standard coal price is 700 yuan / ton, the savings would be 700 yuan / ton × (839808 - 835317) tons = 3.14 million yuan. Attached Figure Description

[0007] Figure 1 A schematic diagram of the combustion optimization control system; Figure 2 This is a schematic diagram of the algorithm in the combustion optimization control system; Figure 3 This is a schematic diagram of the nonlinear rolling optimization module. Figure 4 Deployment architecture diagram for system on-site applications. Detailed Implementation

[0008] The present invention will be further described below with reference to the accompanying drawings and embodiments. The embodiments of the present invention include, but are not limited to, the following embodiments.

[0009] It should be noted that when a component is described as "fixed to" another component, it can be directly on the other component or may have a component in between. When a component is described as "connected to" another component, it can be directly connected to the other component or may have a component in between. When a component is described as "set on" another component, it can be directly set on the other component or may have a component in between. The terms "vertical," "horizontal," "left," "right," "top," "bottom," and similar expressions used in this document are for illustrative purposes only.

[0010] like Figure 1 As shown, the closed-loop intelligent optimization control system for the boiler combustion process comprises four modules: a system calculation module, a system monitoring module, a system communication module, and a system interface module. These modules are independent of each other, but are interconnected through data transfer.

[0011] The functions of each module are as follows: (a) System monitoring module The system monitoring module monitors the operating status of each module at regular intervals and promptly corrects errors in malfunctioning modules using a defined method. If correction fails, the system monitoring module will exit the combustion optimization control algorithm according to a predetermined strategy and smoothly switch the control mode to the original control system in the coal-fired power plant's DCS system, thereby ensuring the normal operation of the coal-fired power plant. Before startup, the system monitoring module needs to perform corresponding checks and judgments. Once the startup conditions are met, it calls and runs the entire control program and periodically checks the operating status of each process, taking appropriate action against malfunctioning modules. (b) System computing module The system calculation module is responsible for calculating optimized parameters such as secondary air, combustion air, coal feed offset, and oxygen setpoint for each layer of the furnace. It also updates the online support vector machine model if the modeling error is significant. This module is the core technology of the combustion optimization control system. In each control cycle, the system calculation module determines whether the current optimization system is in operation. If so, it calculates the output control parameters according to the process; otherwise, it tracks the output control parameters of the original control system.

[0012] (c) System communication module The system communication module primarily enables data communication between the combustion optimization software and the DCS system, comprising two parts: reading data and writing data. The reading part writes data from the DCS system to shared memory for use by other system modules, while the writing part transmits data from the shared memory to the DCS system.

[0013] (d) User interface module The system interface module displays relevant data in real time, allowing operators to understand the system's operating status. It also allows staff to modify and adjust relevant parameters, enabling visualized operation of the combustion optimization software.

[0014] like Figure 2 As shown, the system's computing module is structured as follows: This includes a nonlinear dynamic model and adaptive update module, as well as a nonlinear rolling optimization module. In each control cycle, the nonlinear dynamic model first calculates the boiler efficiency Y at the current moment based on the load and other input parameters. m (k) and NO x The model predicts values ​​and compares them with measured values. Based on the magnitude of the prediction deviation, it is determined whether the model accuracy meets the requirements. If the requirements are not met, the adaptive update module is activated to update the model online. Then, the model is further calibrated through a feedback correction process. Finally, the model is fed into the nonlinear rolling optimization module, which solves the constrained nonlinear optimization problem online to obtain and output the optimal values ​​of the corresponding control variables such as oxygen setpoint, combustion damper opening, secondary damper opening, and coal feed offset.

[0015] A nonlinear dynamic model is established based on various operational variables (such as oxygen setpoint bias, OFA damper opening, secondary damper opening, coal feeder bias, etc.) and various optimization variables (such as fly ash carbon content, air preheater outlet temperature, SCR inlet NOx concentration, CO concentration, superheated steam temperature, reheated steam temperature, economizer outlet flue gas temperature, etc.). A dynamic model of the boiler was established using an online least squares support vector machine. Training samples were organized according to the model's input-output structure, based on historical field operation data and combustion optimization adjustment test data, thus establishing the initial model data. The online adaptive least squares support vector machine is an improvement on the basic least squares support vector machine; when the model changes, the established model can be corrected by timely updating the model library and decision function coefficients.

[0016] For time-varying furnace combustion dynamic systems, the relationship between inputs and outputs is often unpredictable. When using online support vector machines for data-driven "black box" modeling, observing only the relationship between inputs and outputs within a certain time interval results in a steady-state model, but the predictions are often inaccurate. The inputs to a furnace combustion dynamic model should include not only the actual system's input at the current sampling time but also the system's inputs and outputs at previous sampling times.

[0017] The boiler combustion dynamic model established in this application considers total coal feed, burnout air, secondary air, and oxygen content as the main factors affecting boiler emissions and efficiency. Based on the on-site control logic, the same control variable is used for each layer of secondary air commands and burnout air commands. To simplify the model structure, six burner windbox damper commands are employed. U seca This indicates the impact of secondary winds at level 6; the O2 content in the flue gas is used. U O2 Indicates the impact of flue gas O2 content; 1 combustor damper control command. U sofa This indicates the effect of 4 layers of burning wind. The model output is NO. x Emissions and boiler efficiency value Y m (k). Based on the actual dynamic characteristics of the boiler on site, the order of the influence of the input variables on the output variables and the order of the output variables themselves were considered during the dynamic modeling process. Therefore, the input end is increased by adding the current time and previous time NO. x Emissions and boiler efficiency value Y m (k).

[0018] The nonlinear rolling optimization module adopts a multi-objective predictive control method to directly optimize economic indicators (boiler efficiency), NOx emissions, and dynamic performance indicators (superheated steam temperature deviation and reheated steam temperature deviation). It also considers measurable disturbances such as unit load and main steam pressure, unmeasurable disturbances such as coal quality changes, and SCR operating costs to finally obtain the control quantity and achieve closed-loop, dynamic combustion optimization.

[0019] It also includes an online boiler efficiency calculation module for calculating economic indicators. Boiler efficiency is used as the criterion for evaluating boiler economics. Since efficiency is generally not directly measurable, an inverse balancing method is often used to derive it. The system adopts the American Society of Mechanical Engineers' Power Plant Performance Test Code (ASMEPTC). In actual boiler operation, due to technical limitations and instrumentation costs, comprehensive real-time measurement of relevant parameters in the ASME standard is difficult. To facilitate online calculation, a simplified algorithm is used to evaluate boiler operating conditions.

[0020] This application utilizes an external intelligent controller on top of a traditional DCS system. It employs a more stable CPU controller and a superior operating system, without increasing the DCS processor load, and offers excellent real-time performance. Leveraging the controller's powerful computing capabilities, advanced functional modules are developed to solve challenges such as nonlinear, time-varying, and large-delay control, as well as key technical issues like mechanistic model predictive control, which are unsolvable by conventional control methods. This enables intelligent control of the unit operation. Advanced control algorithms are employed, allowing parameters to rapidly adapt to changes in unit load and coal type. The control system exhibits excellent regulation quality under various operating conditions, reducing overshoot and transient time during system regulation, effectively ensuring the unit's economic efficiency and safety. Furthermore, the external intelligent control system is fully functional and highly editable. Modifications to control logic / parameters, functional expansions, and software loading can be achieved simply by disconnecting the external system.

[0021] This application provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the steps of the above-described method.

[0022] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the methods described above. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, or optical storage, etc. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc.

[0023] In the description of this specification, the references to terms such as "some embodiments," "other embodiments," "ideal embodiments," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example that are included in at least one embodiment or example of this application. In this specification, the illustrative descriptions of the above terms do not necessarily refer to the same embodiments or examples.

[0024] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0025] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.

Claims

1. A boiler combustion process closed loop intelligent optimization control system, comprising four modules: system calculation module, system monitoring module, system communication module and system interface module; characterized in that: The system calculation module calculates the optimized quantity of the secondary air, the combustion air, the coal supply bias and the oxygen quantity setting of each layer of the furnace, and simultaneously, if the modeling error is large, the online support vector machine model is updated; in each control period, the system calculation module needs to judge whether the current optimization system is put into operation, if yes, the output control quantity is calculated according to the flow, otherwise, the output control quantity of the original control system is tracked. 2.The closed-loop intelligent optimization control system of the boiler combustion process according to claim 1, characterized in that, The system calculation module comprises a nonlinear dynamic model and an adaptive updating module, and a nonlinear rolling optimization module; In each control cycle, the nonlinear dynamic model first calculates the current boiler efficiency Y according to the load and other input parameters m (k) and NO x The predicted value is compared with the measured value, and then the model accuracy is judged according to the prediction deviation. If the model accuracy does not meet the requirements, the adaptive updating module is started to update the model based on the online updating strategy. Then the model is further calibrated through the feedback correction link. Finally, the nonlinear rolling optimization module is sent in, and the corresponding control variables of the oxygen amount set value, the combustion air damper opening, the secondary air damper opening and the coal supply amount bias are obtained by online solving the constrained nonlinear optimization problem. 3.The closed-loop intelligent optimization control system of the boiler combustion process according to claim 2, characterized in that, The boiler dynamic model is established by using an online least square support vector machine based on each operation variable and each optimized variable. 4.The closed-loop intelligent optimization control system of the boiler combustion process according to claim 3, characterized in that, When the boiler dynamic model is established by using the online least square support vector machine, the initial model data is established by organizing training samples from the field operation history data and combining with the combustion optimization adjustment test data according to the model input and output structure; the online adaptive least square support vector machine is improved based on the least square support vector machine, and when the model changes, the built model is corrected by timely updating the model library and the decision function coefficient. 5.The closed-loop intelligent optimization control system of the boiler combustion process according to claim 2, characterized in that, The boiler combustion dynamic model introduces the total coal amount, overfire air, secondary air, and oxygen content as the main factors affecting the boiler emission and efficiency. According to the field control logic, the same control quantity is used for the secondary air command and overfire air command of each layer, and the 6 burner damper commands are used U seca represents the influence of 6 layers of secondary air; the flue gas O2 content is used U O2 represents the influence of the flue gas O2 content; 1 overfire air damper control command U sofa represents the influence of 4 layers of overfire air; the model output is the NOx emission and the boiler efficiency value Ym(k). 6.The closed-loop intelligent optimization control system of the boiler combustion process according to claim 2, characterized in that, The nonlinear rolling optimization module adopts a multi-objective predictive control method to optimize the economic index, the NOx emission and the dynamic performance index, simultaneously introduces the measurable disturbance of the unit load and the main steam pressure, the unmeasurable disturbance of the coal quality change and the SCR operation cost factor, and finally obtains the control quantity to realize the closed-loop and dynamic combustion optimization. 7.The closed-loop intelligent optimization control system of the boiler combustion process according to claim 6, characterized in that, The boiler efficiency online calculation module is used to calculate the economic index, adopts a counterbalance method to obtain the boiler efficiency, and the algorithm comes from the American Society of Mechanical Engineers' power station performance test regulation. 8.The closed-loop intelligent optimization control system of the boiler combustion process according to claim 1, characterized in that, The system monitoring module monitors the running state of each module at a certain period, timely corrects the error of the module in a certain way, and if the correction cannot be realized, the system monitoring module exits the combustion optimization control algorithm according to the established strategy, and smoothly switches the control mode to the original control system in the coal-fired power station DCS system. 9.The closed-loop intelligent optimization control system of the boiler combustion process according to claim 1, characterized in that, The system communication module realizes the data communication between the DCS system, including reading data and writing data. 10.The closed-loop intelligent optimization control system of the boiler combustion process according to claim 1, characterized in that, The system interface module has the functions of displaying relevant data in real time, facilitating the operation personnel to understand the running state of the system, allowing the operation personnel to modify and adjust relevant parameters, and realizing visual operation of the combustion optimization software.

Images