WO2023087906A1 - 一种基于事件触发的模糊神经网络温度控制系统及方法 - Google Patents

一种基于事件触发的模糊神经网络温度控制系统及方法 Download PDF

Info

Publication number
WO2023087906A1
WO2023087906A1 PCT/CN2022/120450 CN2022120450W WO2023087906A1 WO 2023087906 A1 WO2023087906 A1 WO 2023087906A1 CN 2022120450 W CN2022120450 W CN 2022120450W WO 2023087906 A1 WO2023087906 A1 WO 2023087906A1
Authority
WO
WIPO (PCT)
Prior art keywords
neural network
temperature
loop control
fuzzy
controller
Prior art date
Application number
PCT/CN2022/120450
Other languages
English (en)
French (fr)
Inventor
杜昭平
方雨帆
杨晓飞
李建祯
邹治林
沈帅
Original Assignee
江苏科技大学
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 江苏科技大学 filed Critical 江苏科技大学
Publication of WO2023087906A1 publication Critical patent/WO2023087906A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
    • G05D23/00—Control of temperature
    • G05D23/19—Control of temperature characterised by the use of electric means
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/043—Architecture, e.g. interconnection topology based on fuzzy logic, fuzzy membership or fuzzy inference, e.g. adaptive neuro-fuzzy inference systems [ANFIS]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
    • Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
    • Y02P90/02—Total factory control, e.g. smart factories, flexible manufacturing systems [FMS] or integrated manufacturing systems [IMS]

Definitions

  • the invention relates to the technical field of thermal power generation temperature control, in particular to a thermal power plant main steam temperature control system and control method based on event-triggered fuzzy RBF neural network PID control.
  • the main steam temperature is a key parameter for thermal process control of thermal power plant boilers. According to the experience of the operators of thermal power plants, when the load disturbance of the unit is relatively large, improper operation by the operator will easily cause accidents, and in severe cases, the superheater will be overheated, and the unit may even be shut down due to superheater leakage, seriously affecting the operation of the unit. safety and stability.
  • the main steam temperature of the drum boiler usually adopts the conventional cascade control system, which has large inertia, delay, and nonlinearity, so an improved control strategy is proposed.
  • the main steam temperature is too high or too low due to the following factors: mainly related to the main steam flow rate, the layout structure of the tail flue superheater, and the type of superheater, heat transfer method, flue gas flow rate, heat transfer method and other factors; from From the perspective of unit operation control process: the main steam temperature is overheated or too low, which is mainly caused by unfavorable monitoring parameters and improper operation of the operating personnel; from the perspective of the structure of the main steam temperature control system, it is largely due to improper parameter setting in the design caused by. Poor control of the main steam temperature not only affects the safety and stable operation of the unit, but also has a significant impact on the life of the corresponding equipment of the unit, especially the superheater and the gas turbine.
  • the main steam temperature is a monitoring parameter for thermal power operators. It is necessary to set up a separate operation specialist post to control the main steam temperature. Similarly, thermal technicians analyze the deficiencies in the control system according to the main steam temperature control curve in daily operation and maintenance. Then put forward a perfect control strategy or need to redesign.
  • the excellent quality of boiler main steam temperature is an essential performance of modern large-capacity and high-parameter thermal power units. It runs through the integration process of the entire control system from design, installation, commissioning, testing, and operation, as well as whether the corresponding DCS system platform is perfect. ; Therefore, a safe, stable and effective boiler main steam temperature control system is very important for thermal power boiler superheater equipment and steam turbine operation.
  • the conventional main steam temperature control in thermal power plants generally combines strategies such as feedforward compensation and cascade control systems, and the design method of the cascade control system is: the main and auxiliary controllers adopt PID controllers.
  • the three parameters of proportional, integral, and differential must be adjusted first, not only the tuning process of the main loop PID parameters, but also the secondary loop parameters
  • the proportional, integral and differential parameters basically do not change, but when the working condition of the machine unit changes, the proportional, integral and differential parameters are no longer suitable for the control needs, and offline tuning is required.
  • the secondary controller receives the status signal output by the desuperheater and the output signal of the main controller. When the superheated temperature rises, the output of the main controller decreases, the output of the secondary controller increases, the amount of desuperheating water increases, and the superheated temperature decreases.
  • the above-mentioned cascade control system has two loops, inner and outer, and the outer loop is composed of the main steam object, the main transmitter, the state observer, the main controller and the entire inner loop.
  • the auxiliary circuit includes auxiliary detection transmitter, auxiliary controller, actuator, desuperheating water valve, desuperheater, superheater, etc.
  • the inner loop is a follow-up control system, and the secondary loop needs the output of the main controller of the outer loop as the set value, and uses the output of the secondary controller to control the action of the actuator to realize the control of the desuperheater. Because the delay and inertia of the secondary loop are small, its control process is stable.
  • the system can adjust in time to quickly and stably reduce the disturbance, especially the influence of the desuperheater water disturbance on the superheated steam temperature ;
  • the outer loop is a low-speed loop, and its main task is to maintain the main steam temperature equal to a given value.
  • the main steam temperature has complex dynamic and strong coupling characteristics.
  • the conventional PID control described above only focuses on the relationship between a single input and output variable in the control loop, but cannot compensate the relationship between strongly coupled or sub-strongly coupled input and output variables.
  • the invention provides an event-triggered fuzzy neural network temperature control system and method to solve the problem in the prior art that the main steam temperature is not ideally controlled under variable loads and affects the safety of the unit.
  • the present invention provides an event-triggered fuzzy neural network temperature control system, comprising: an outer loop control loop and an inner loop control loop, wherein the outer loop control loop and the inner loop control loop form a cascade control loop;
  • the outer loop control loop includes: a main controller, a main steam object, and a main transmitter;
  • Described master controller comprises: event trigger, fuzzy RBF neural network, PID controller;
  • the inner control loop includes: a secondary controller, an actuator, a desuperheater, and a secondary transmitter;
  • the output end of the main steam object is connected with the input end of the main transmitter; the output end of the main transmitter is connected with the input end of the event trigger; the output end of the event trigger is connected with The input terminal of the fuzzy RBF neural network is connected; the output terminal of the fuzzy RBF neural network is connected with the input terminal of the PID controller; the output terminal of the PID controller is connected with the input terminal of the secondary controller; the secondary The output end of the controller is connected to the input end of the actuator; the output end of the actuator is connected to the input end of the desuperheater; the output end of the desuperheater is respectively connected to the input end of the main steam object connected to the input terminal of the secondary transmitter; the output terminal of the secondary transmitter is connected to the input terminal of the secondary controller.
  • the outer loop control loop also includes: a state observer
  • the output end of the main transmitter is connected to the input end of the state observer; the output end of the state observer is respectively connected to the input end of the event trigger and the input end of the secondary controller.
  • main controller also includes: a domain adjuster, a neural network parameter adjuster;
  • the output end of the state observer is connected with the input end of the universe adjuster; the output end of the universe adjuster is connected with the input end of the fuzzy RBF neural network; the neural network parameter adjuster is connected with the The fuzzy RBF neural network connection is described.
  • the present invention also provides a control method based on an event-triggered fuzzy neural network temperature control system, including: an outer loop control loop control method, and an inner loop control loop control method; wherein:
  • Step A1 Collect the temperature of the main steam object through the main transmitter to obtain the main steam temperature signal
  • Step A2 Comparing the main steam temperature signal with the standard temperature signal, calculating the temperature deviation and the temperature deviation change rate;
  • Step A3 The event trigger judges the event trigger according to the temperature deviation change rate. When the event is triggered, the event trigger outputs the received temperature deviation change rate; when the event is not triggered, the event trigger does not output a signal;
  • Step A4 When the fuzzy RBF neural network receives the rate of change of temperature deviation, adjust the three parameters of the PID controller according to the rules of the fuzzy RBF neural network until the parameters of the PID controller reach the optimum;
  • Step A5 The PID controller outputs the outer loop control signal according to the control parameters to complete the outer loop control loop control
  • Step B1 Collect the temperature of the desuperheater through the auxiliary transmitter to obtain the temperature signal of the desuperheater;
  • Step B2 the sub-controller generates an inner loop control signal according to the outer loop control signal and the temperature signal of the desuperheater;
  • Step B3 The actuator acts according to the inner loop control signal to complete the control of the inner loop control loop.
  • step A2 also includes: the state observer generates a temperature feedback compensation signal according to the main steam temperature signal;
  • the step B2 the sub-controller generates an inner loop control signal according to the outer loop control signal, the temperature feedback compensation signal and the desuperheater temperature signal.
  • step A3 also includes: adjusting the expansion factor in the fuzzy RBF neural network according to the temperature deviation amount and the temperature deviation change rate by the universe adjuster;
  • the neural network parameter adjuster adjusts the connection weight, membership function center and base width in the fuzzy RBF neural network.
  • the trigger condition of the event trigger is:
  • de/dt((k+i)h) is the temperature deviation change rate at the current moment (k+i)
  • de/dt(kh) is the temperature deviation change rate at the previous moment (k)
  • the action of the actuator according to the control signal of the inner loop is as follows: the actuator adjusts the desuperheating valve installed on the desuperheating water pipeline connected between the actuator and the desuperheater, and adjusts the amount of water flowing into the desuperheating water pipeline. The flow rate of desuperheating water.
  • the secondary controller adopts PI control mode.
  • the sub-controller of cascade control adopts traditional PI control, and the main controller adopts PID controller based on event trigger and fuzzy RBF neural network, which can obtain system The temperature deviation and deviation change rate, and then input these two parameters into the fuzzy RBF neural network, through the fuzzy RBF neural network regularizer, online self-tuning of the three parameters of PID, and finally achieve the most ideal control Effect;
  • the event trigger is used to judge the output value according to the synchronization signal received at the current moment and its internal event trigger mechanism rules, and through the latest output value of the event trigger and the latest The comparison between the received values is used to determine the output value of the main controller, which can reduce the fluctuation range of the main steam temperature, improve the adjustment quality, and reduce the adjustment times of the adjustment valve, thereby improving the service life.
  • Fig. 1 is a structural schematic diagram of a cascade temperature control system of a traditional cascade thermal power plant
  • Fig. 2 is a structural representation of the present invention
  • Fig. 3 is a schematic diagram of the change of the fuzzy universe of universe adjuster of the present invention.
  • the embodiment of the present invention provides an event-triggered fuzzy neural network temperature control system, as shown in Figure 2, including: an outer loop control loop, an inner loop control loop, the outer loop control loop and the inner loop control loop constitute a cascade control loop ;
  • the outer loop control loop includes: main controller, main steam object, main transmitter;
  • the main controller includes: event trigger, fuzzy RBF neural network, PID controller;
  • the inner loop control loop includes: sub-controller, actuator, desuperheater, sub-transmitter;
  • the output end of the main steam object is connected with the input end of the main transmitter; the output end of the main transmitter is connected with the input end of the event trigger; the output end of the event trigger is connected with the input end of the fuzzy RBF neural network; fuzzy The output of the RBF neural network is connected to the input of the PID controller; the output of the PID controller is connected to the input of the sub-controller; the output of the sub-controller is connected to the input of the actuator; the output of the actuator is connected to the reducer
  • the input end of the desuperheater is connected to the input end of the desuperheater; the output end of the desuperheater is respectively connected to the input end of the main steam object and the input end of the auxiliary transmitter; the output end of the auxiliary transmitter is connected to the input end of the auxiliary controller.
  • the main steam temperature signal X1(t) is time sampled by the main transmitter to obtain the temperature sampling signal X1(kh), and the obtained main steam temperature signal X1(kh) is compared with the standard temperature signal r(kh) in the first comparator Finally, the temperature deviation change e is obtained, and the temperature deviation change e is passed through the differentiator to obtain the temperature deviation change rate ec as the input of the event trigger; the temperature deviation change ec output after filtering is output to the fuzzy RBF neural network, and the neural network
  • the parameter adjuster works and outputs the optimal parameters of the PID controller; the PID controller outputs the control signal u1(kh) as the input terminal of the sub-controller, and the control signal u2(kh) output by the sub-controller controls the action of the actuator and executes
  • the desuperheater controls the work of the desuperheater to adjust the steam temperature; the temperature signal X2(t) output by the desuperheater is time-sampled by
  • the outer loop control loop also includes: a state observer
  • the output end of the main transmitter is connected with the input end of the state observer; the output end of the state observer is respectively connected with the input end of the first comparator and the input end of the secondary controller.
  • the output temperature deviation signal e of the first comparator is connected to the input terminal of the differentiator, and the temperature deviation change rate signal ec outputted by the temperature deviation variation after passing through the differentiator is used as the input terminal of the event trigger.
  • the output terminal of the state observer is connected to the second comparator, and the output terminal of the second comparator is connected to another input terminal of the state observer.
  • b0 is a special parameter, which plays a role in compensating the control quantity.
  • the input of the state observer is the temperature sampling signal X1(kh), the output signal Z2(kh) of the second comparator, the output compensation signal Z1(kh) of the state observer is used as the input terminal of the second comparator, and the main steam temperature signal Z3(kh) is used as the input terminal of the first comparator to compare with the standard temperature signal r(kh) to obtain the temperature deviation signal e, the output temperature deviation signal e of the first comparator is connected to the input terminal of the differentiator, and the temperature deviation The temperature deviation change rate signal ec output by the differentiator after the variation e is used as the input terminal of the event trigger.
  • the main controller also includes: domain adjuster, neural network parameter adjuster;
  • the output end of the first comparator and the output end of the differentiator are connected with the input end of the universe adjuster; the output end of the universe adjuster is connected with the input end of the fuzzy RBF neural network; the neural network parameter adjuster is connected with the fuzzy RBF neural network connect.
  • the two input terminals of the domain adjuster are the temperature deviation change amount and the temperature deviation change rate, and the universe adjuster can adjust the expansion factor ⁇ according to the values of these two quantities, where ⁇ [0,1 ], the domain of discourse adjustment can reduce the domain of discourse to [- ⁇ E, ⁇ E] through the expansion factor ⁇ .
  • the number of fuzzy variables remains unchanged, the division of fuzzy variables on the unit domain of discourse near the zero point is intensive, which is equivalent to indirectly increasing the fuzzy control rules , to improve the sensitivity of the control.
  • the domain of discourse can be expanded to (- ⁇ E, ⁇ E) by the expansion factor ⁇ (1, ⁇ ), which is conducive to accelerating the system response and reducing the adjustment time , to obtain excellent control performance in all working conditions;
  • the neural network parameter adjuster adjusts the connection weight, membership function center and base width in the fuzzy RBF neural network.
  • the present invention also provides a control method based on an event-triggered fuzzy neural network temperature control system, including: an outer loop control loop control method, and an inner loop control loop control method; wherein:
  • Step A1 Collect the temperature of the main steam object through the main transmitter to obtain the main steam temperature signal
  • Step A2 Comparing the main steam temperature signal with the standard temperature signal, calculating the temperature deviation and temperature deviation change rate; the state observer generates a temperature feedback compensation signal according to the main steam temperature signal;
  • Step A3 The event trigger judges the event trigger according to the temperature deviation change rate. When the event is triggered, the event trigger outputs the received temperature deviation change rate; when the event is not triggered, the event trigger does not output a signal;
  • the triggering conditions of the event trigger are:
  • de/dt((k+i)h) is the temperature deviation change rate at the current moment (k+i)
  • de/dt(kh) is the temperature deviation change rate at the previous moment (k)
  • the basic design idea of the event trigger rule is: calculate the differential deviation signal value received at the current moment and the differential deviation signal value received at the previous moment, and compare the two values for difference.
  • the newly received signal will not be transmitted; in this example, if the trigger function satisfies the condition less than or equal to ⁇ , it is considered that no "event" has occurred, and the newly received signal de/dt(k +i) will not be output. If the trigger function satisfies the condition greater than ⁇ , it is considered that an "event" has occurred, and the event trigger will output the newly accepted signal de/dt(k+i) to the fuzzy RBF neural network for Update the three parameters of the PID controller output by the fuzzy RBF neural network, adjust the control valve action of the controlled process, and realize the control of the entire system;
  • Step A4 When the fuzzy RBF neural network receives the rate of change of temperature deviation, adjust the three parameters of the PID controller according to the rules of the fuzzy RBF neural network until the parameters of the PID controller reach the optimum;
  • Step A5 The PID controller outputs the outer loop control signal according to the control parameters to complete the outer loop control loop control
  • the PID controller uses the Kp, Ki, and Kd parameter values output by the fuzzy RBF neural network as the parameter values of the PID controller.
  • the online self-tuning process of the three parameters of the PID by the fuzzy RBF neural network is as follows:
  • Step B1 Collect the temperature of the desuperheater through the auxiliary transmitter to obtain the temperature signal of the desuperheater;
  • Step B2 The sub-controller generates an inner loop control signal according to the outer loop control signal, the temperature feedback compensation signal and the desuperheater temperature signal;
  • Step B3 The actuator acts according to the inner loop control signal to complete the control of the inner loop control loop.
  • Step A3 also includes: adjusting the scaling factor in the fuzzy RBF neural network according to the temperature deviation amount and temperature deviation change rate by the universe adjuster;
  • the neural network parameter adjuster adjusts the connection weight, membership function center and base width in the fuzzy RBF neural network.
  • the fuzzy RBF neural network is a fuzzy control algorithm realized by the neural network structure.
  • the neural network parameter adjuster can be used to determine and change the connection weight wij, the membership function center cij and the base width bij of the fuzzy RBF neural network.
  • the output of the controller is :
  • e(k) is the systematic deviation at the kth sampling moment
  • u(k) is the output value at the kth sampling moment
  • ⁇ u(k) is the output increment at the kth sampling moment
  • the system uses a supervised learning algorithm, and the objective function of learning is defined as:
  • r(k) and y(k) are the ideal output and actual output of the system at time kT respectively, and r(k)-y(k) represents the control error of iteration step k;
  • the above control method uses the gradient descent method to search and optimize, and the iterative algorithm of output weight wij, membership function center cij and base width bij is:
  • step B3 the action of the actuator according to the control signal of the inner loop is as follows: the actuator adjusts the desuperheating valve installed on the desuperheating water pipeline connected between the actuator and the desuperheater, and adjusts the flow rate of the desuperheating water flowing into the desuperheating water pipeline .

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Software Systems (AREA)
  • General Health & Medical Sciences (AREA)
  • Data Mining & Analysis (AREA)
  • Automation & Control Theory (AREA)
  • Mathematical Physics (AREA)
  • General Engineering & Computer Science (AREA)
  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Artificial Intelligence (AREA)
  • Biomedical Technology (AREA)
  • Biophysics (AREA)
  • Computational Linguistics (AREA)
  • Computing Systems (AREA)
  • Evolutionary Computation (AREA)
  • Molecular Biology (AREA)
  • Computational Mathematics (AREA)
  • Mathematical Analysis (AREA)
  • Fuzzy Systems (AREA)
  • Pure & Applied Mathematics (AREA)
  • Mathematical Optimization (AREA)
  • Feedback Control In General (AREA)
  • Control Of Temperature (AREA)

Abstract

一种基于事件触发的模糊神经网络温度控制系统，包括：主蒸汽对象的输出端与主变送器的输入端连接；主变送器的输出端与事件触发器的输入端连接；事件触发器的输出端与模糊RBF神经网络的输入端连接；模糊RBF神经网络的输出端与PID控制器的输入端连接；PID控制器的输出端与副控制器的输入端连接；副控制器的输出端与执行器的输入端连接；执行器的输出端与减温器的输入端连接；减温器的输出端分别与主蒸汽对象的输入端、副变送器的输入端连接；副变送器的输出端与副控制器的输入端连接。

Description

一种基于事件触发的模糊神经网络温度控制系统及方法 技术领域
本发明涉及火电发电温度控制技术领域，具体涉及计一种基于事件触发的模糊RBF神经网络PID控制的火电厂主蒸汽温度控制系统和控制方法。
背景技术
主蒸汽温度是火电厂锅炉热工过程控制的关键参数。根据火电厂运行人员的经验,当机组负荷扰动比较大时,运行人员操作不当很容易造成事故的发生,严重时导致过热器超温,甚至可能出现过热器漏泄而使机组停机,严重影响机组运行的安全和稳定。汽包锅炉主蒸汽温度通常采用常规串级控制系统存在大惯性、延迟性、非线性而提出改进控制策略的设想。主蒸汽温度过高或者过低主下列因素:主要有主蒸汽流量大小、尾部烟道过热器布置结构,以及过热器的类型、换热方式、烟气流量、传热方式等因素有关系；从机组运行控制过程来看:主蒸汽温度超温或者过低主要由运行人员监视参数不利和操作不当造成的；从主蒸汽温度控制系统结构来看,很大程度上是由于设计上存在参数整定不当所引起的。主蒸汽温度控制不好不但影响机组的安全和稳定运行,而且对机组相应的设备尤其是过热器和气轮机的寿命影响重大,尤其温度低不仅会损坏气轮机末级叶片,严重发生水击现象。因此,主蒸汽温度是火电运行人员监视参数需要单独设置一个运行专员岗位来对主蒸汽温度进行控制,同样热工技术人员在日常行和维护中根据主蒸汽温度控制曲线来分析控制系统存在的不足进而提出完善的控制策略或者需要重新进行设计。锅炉主蒸汽温度优良的品质是现代大容量、高参数火电机组必备的性能,从设计、安装、调试、检测、运行等环节都贯穿于整个控制系统的集成过程以及对应的DCS系统平台是否完善；因此,安全、稳定、有效的锅炉主蒸汽温度控制系统对火电锅炉过热器设备和汽轮机运行非常重要。
目前火电厂常规的主蒸汽温度控制一般结合前馈补偿和串级控制系统等策略，且串级控制系统的设计方法是：其主、副控制器采PID控制器。通常采用的 比例、积分、微分主蒸汽温度串级控制系统,在投入运行之前,首先要对比例、积分、微分三个参数进行整定,不仅有主回路PID参数的整定过程,还有副回路参数的整定过程；当投入运行之后,比例、积分、微分参数基本不在改变,但当机机组工况发生变化时,比例、积分、微分参数不在适合控制的需要,需要离线整定。副控制器接受减温器输出的状态信号和主控制器输出信号。当过热气温升高时，主控制器输出减小，副控制器输出增加，减温水量增加，过热气温下降。
如图1所示，上述串级控制系统中具有内、外两个回路，外回路由主蒸汽对象、主变送器、状态观测器、主控制器以及整个内回路构成的。副回路包括副检测变送器、副控制器、执行器、减温水阀门、减温器、过热器等。此外内回路还是一个随动控制系统，副回路需要以外回路主控制器的输出为设定值，并利用副控制器的输出来控制执行器动作，实现对减温器的控制。因为副回路迟延和惯性较小，因此它的控制过程是稳定的。当减温水发生扰动时或减温器后的过热器出口蒸汽温度发生变化而引起导前汽温变化时，系统能及时调整，快速稳定减小扰动、特别是减温水扰动对过热汽温的影响；相对于内回路，外回路是一个低速回路，它的主要任务是维持主汽温等于给定值。主蒸汽温度有着复杂的动态和强耦合特性。上面所述常规的PID控制仅仅关注控制回路中单个输入输出变量之间的关系，而无法对强耦合或者次强耦合的输入输出变量之间的关系予以补偿。在实际运行中，一方面由于副控制器的不断调节，使得控制阀等执行器频繁操作，降低了使用寿命；另一方面，这种常规的主汽温控制策略，采用固定参数或分段PID构造控制器，没有完全考虑主汽温在变负荷下模型变化的影响，控制效果仍会很不理想，严重影响了机组的经济性和安全性。
发明内容
本发明提供了一种基于事件触发的模糊神经网络温度控制系统及方法，以解决现有技术中主汽温在变负荷下控制不理想，影响机组安全性的问题。
本发明提供了一种基于事件触发的模糊神经网络温度控制系统，包括：外环控制回路、内环控制回路，所述外环控制回路与内环控制回路构成串级控制回路；
所述外环控制回路包括：主控制器、主蒸汽对象、主变送器；
所述主控制器包括：事件触发器、模糊RBF神经网络、PID控制器；
所述内环控制回路包括：副控制器、执行器、减温器、副变送器；
所述主蒸汽对象的输出端与所述主变送器的输入端连接；所述主变送器的输出端经与所述事件触发器的输入端连接；所述事件触发器的输出端与模糊RBF神经网络的输入端连接；所述模糊RBF神经网络输出端与所述PID控制器的输入端连接；所述PID控制器的输出端与所述副控制器的输入端连接；所述副控制器的输出端与所述执行器的输入端连接；所述执行器的输出端与所述减温器的输入端连接；所述减温器的输出端分别与所述主蒸汽对象的输入端、副变送器的输入端连接；所述副变送器的输出端与所述副控制器的输入端连接。
进一步地，所述外环控制回路还包括：状态观测器；
所述主变送器的输出端与所述状态观测器的输入端连接；所述状态观测器的输出端分别与所述事件触发器的输入端、副控制器的输入端连接。
进一步地，所述主控制器还包括：论域调整器、神经网络参数调整器；
所述状态观测器的输出端与所述论域调整器的输入端连接；所述论域调整器的输出端与所述模糊RBF神经网络的输入端连接；所述神经网络参数调整器与所述模糊RBF神经网络连接。
本发明还提供了一种基于事件触发的模糊神经网络温度控制系统的控制方法，包括：外环控制回路控制方法、内环控制回路控制方法；其中：
所述外环控制回路控制方法步骤如下：
步骤A1：通过主变送器对主蒸汽对象进行温度采集，获取主蒸汽温度信号；
步骤A2：将主蒸汽温度信号与标准温度信号进行比较，计算温度偏差量、温度偏差变化率；
步骤A3：事件触发器根据温度偏差变化率进行事件触发判断，当事件触发时，事件触发器输出接收的温度偏差变化率；当事件不触发时，事件触发器不输出信号；
步骤A4：当模糊RBF神经网络接收到温度偏差变化率时，根据模糊RBF神经网络规则对PID控制器的三个参数进行整定，直至PID控制器的参数达到最优；
步骤A5：PID控制器根据控制参数输出外环回路控制信号，完成外环控制 回路控制；
所述内环控制回路控制方法步骤如下：
步骤B1：通过副变送器对减温器进行温度采集，获取减温器温度信号；
步骤B2：副控制器根据所述外环回路控制信号以及减温器温度信号产生内环控制信号；
步骤B3：执行器根据内环控制信号动作，完成内环控制回路控制。
进一步地，所述步骤A2中还包括：状态观测器根据主蒸汽温度信号产生温度反馈补偿信号；
所述步骤B2：副控制器根据所述外环回路控制信号、温度反馈补偿信号以及减温器温度信号产生内环控制信号。
进一步地，所述步骤A3中还包括：论域调整器根据温度偏差量、温度偏差变化率，调节模糊RBF神经网络中的伸缩因子；
神经网络参数调整器调节模糊RBF神经网络中的连接权、隶属度函数中心和基宽。
进一步地，所述事件触发器的触发条件为：
||de/dt((k+i)h)-de/dt(kh)||≤σ
其中，de/dt((k+i)h)是当前时刻(k+i)的温度偏差变化率，de/dt(kh)是上一时刻(k)的温度偏差变化率，||||表示范数，σ为(0,1)的有界正数，i＝1,2,…,为正整数。
进一步地，所述步骤B3中执行器根据内环控制信号动作具体为：执行器对设置在执行器与减温器相连通的减温水管道上的减温阀进行调节，调节流入减温水管道中的减温水的流量。
进一步地，所述副控制器为PI控制方式。
本发明的有益效果：
(1)串级控制的副控制器采用传统PI控制，主控制器采用基于事件触发器和模糊RBF神经网络的PID控制器，能够根据当前时刻主蒸汽温度的输出和设定值比较，得到系统的温度的偏差和偏差变化率，然后将得到的这两个参数输入到模糊RBF神经网络中，通过模糊RBF神经网络规则器对PID的三个参数进 行在线自我整定，并最终实现最为理想的控制效果；
(2)将事件触发器引入主控制器中，事件触发器用于依据当前时刻接收到的同步信号和其内部的事件触发机制规则来判断输出值，并通过事件触发器最新的输出值与最新的接收值之间的比较来决定接下来主控制器的输出值，可以减少主蒸汽温度的波动幅度，提高调节品质，同时可以减少调节阀门的调节次数，提高了使用寿命。
附图说明
通过参考附图会更加清楚的理解本发明的特征和优点，附图是示意性的而不应理解为对本发明进行任何限制，在附图中：
图1为传统串级火电厂串级温度控制系统的结构示意图；
图2为本发明结构示意图；
图3为本发明论域调整器的模糊论域变化示意图。
具体实施方式
为使本发明实施例的目的、技术方案和优点更加清楚，下面将结合本发明实施例中的附图，对本发明实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例是本发明一部分实施例，而不是全部的实施例。基于本发明中的实施例，本领域技术人员在没有作出创造性劳动前提下所获得的所有其他实施例，都属于本发明保护的范围。
本发明实施例提供一种基于事件触发的模糊神经网络温度控制系统，如图2所示，包括：外环控制回路、内环控制回路，外环控制回路与内环控制回路构成串级控制回路；
外环控制回路包括：主控制器、主蒸汽对象、主变送器；
主控制器包括：事件触发器、模糊RBF神经网络、PID控制器；
内环控制回路包括：副控制器、执行器、减温器、副变送器；
主蒸汽对象的输出端与主变送器的输入端连接；主变送器的输出端经与事件触发器的输入端连接；事件触发器的输出端与模糊RBF神经网络的输入端连接；模糊RBF神经网络输出端与PID控制器的输入端连接；PID控制器的输出 端与副控制器的输入端连接；副控制器的输出端与执行器的输入端连接；执行器的输出端与减温器的输入端连接；减温器的输出端分别与主蒸汽对象的输入端、副变送器的输入端连接；副变送器的输出端与副控制器的输入端连接。
主蒸汽温度信号X1(t)经主变送器时间采样后得到温度采样信号X1(kh)，将得到的主蒸汽温度信号X1(kh)与标准温度信号r(kh)在第一比较器比较后得到温度偏差变化量e，温度的偏差变化量e经微分器后得到温度偏差变化率ec作为事件触发器的输入；经筛选输出的温度偏差变化量ec输出到模糊RBF神经网络后，神经网络参数调整器工作，输出最优的PID控制器的参数；PID控制器输出控制信号u1(kh)作为副控制器的输入端，副控制器输出的控制信号u2(kh)控制执行器动作，执行器控制减温器工作来调节蒸汽温度；减温器输出的温度信号X2(t)经副变送器时间采样后的温度采样信号X2(kh)输出到副控制器，减温器输出的温度信号y2(t)作为主蒸汽温度的输入端调节主蒸汽温度。
外环控制回路还包括：状态观测器；
主变送器的输出端与状态观测器的输入端连接；状态观测器的输出端分别与第一比较器的输入端、副控制器的输入端连接。第一比较器的输出温度偏差量信号e与微分器的输入端相连，温度偏差变化量经微分器后输出的温度偏差变化率信号ec作为事件触发器的输入端。
状态观测器的输出端与第二比较器相连，第二比较器的输出端和状态观测器的另外一个输入端相连，b0是一个特殊的参数，对控制量起到补偿的作用。
状态观测器的输入为温度采样信号X1(kh)、第二比较器的输出信号Z2(kh)，状态观测器的输出补偿信号Z1(kh)作为第二比较器的输入端，主蒸汽温度信号Z3(kh)作为第一比较器的输入端与标准温度信号r(kh)进行比较得到温度偏差量信号e，第一比较器的输出温度偏差量信号e与微分器的输入端相连，温度偏差变化量e经微分器后输出的温度偏差变化率信号ec作为事件触发器的输入端。
主控制器还包括：论域调整器、神经网络参数调整器；
第一比较器的输出端以及微分器的输出端与论域调整器的输入端连接；论域调整器的输出端与模糊RBF神经网络的输入端连接；神经网络参数调整器与模糊RBF神经网络连接。
如图3所示，论域调整器的两个输入端是温度偏差变化量以及温度偏差变化率，论域调整器能够根据这两个量的值调整伸缩因子δ，其中δ∈[0,1]，论域调整能通过伸缩因子δ将论域缩小为[-δE,δE]，虽然模糊变量数量不变，但零点附近单位论域上的模糊变量划分密集，相当于间接增加了模糊控制规则，提高了控制的灵敏度。同理，当初始控制阶段误差e和误差变化率ec较大时，可通过伸缩因子β∈(1,∞)将论域膨胀为(-βE,βE)，有利于加速系统响应，减少调节时间，以获得全工况优良的控制性能；神经网络参数调整器调节模糊RBF神经网络中的连接权、隶属度函数中心和基宽。
本发明还提供了一种基于事件触发的模糊神经网络温度控制系统的控制方法，包括：外环控制回路控制方法、内环控制回路控制方法；其中：
外环控制回路控制方法步骤如下：
步骤A1：通过主变送器对主蒸汽对象进行温度采集，获取主蒸汽温度信号；
步骤A2：将主蒸汽温度信号与标准温度信号进行比较，计算温度偏差量、温度偏差变化率；状态观测器根据主蒸汽温度信号产生温度反馈补偿信号；
步骤A3：事件触发器根据温度偏差变化率进行事件触发判断，当事件触发时，事件触发器输出接收的温度偏差变化率；当事件不触发时，事件触发器不输出信号；
事件触发器的触发条件为：
||de/dt((k+i)h)-de/dt(kh)||≤σ
其中，de/dt((k+i)h)是当前时刻(k+i)的温度偏差变化率，de/dt(kh)是上一时刻(k)的温度偏差变化率，||||表示范数，σ为(0,1)的有界正数，i＝1,2,…,为正整数。事件触发规则的基本设计思路是：计算当前时刻接受的微分偏差信号值与上一时刻接受的微分偏差信号值，将这两个值进行求差比较，若大于设定好的阈值，则认为出发了“事件”，否则新接收的信号就不被传输；在本例中，若触发函数满足小于等于σ的条件，则认为没有发生“事件”，事件触发器新接受的信号de/dt(k+i)不会输出，若触发函数满足大于σ的条件，则认为有发生“事件”，事件触发器会将新接受的信号de/dt(k+i)输出至模糊RBF神经网络，用于更新模糊RBF神经网络输出的PID控制器的三个参数，调节被控过程的控制阀动作，实 现对整个系统控制；
步骤A4：当模糊RBF神经网络接收到温度偏差变化率时，根据模糊RBF神经网络规则对PID控制器的三个参数进行整定，直至PID控制器的参数达到最优；
步骤A5：PID控制器根据控制参数输出外环回路控制信号，完成外环控制回路控制；
PID控制器是根据模糊RBF神经网络输出的Kp、Ki、Kd参数值来作为PID控制器的参数值。模糊RBF神经网络对PID的三个参数进行在线自我整定过程如下：
(1)初始化改控制器中隶属度函数的中心c、基宽b、网络各层系数的初始值w、学习速率η和惯性系数α；
(2)通过获得采样获得系统的实际的输出值y(k)和输入值r(k)，通过计算得出该系统的温度偏差量e(k)以及温度偏差变化率ec(k)；
(3)计算出模糊RBF神经网络中各层神经网络的输入、输出以及PID控制器输出的控制量u1(k)，将u1(k)加入被控对象中，使之产生下一采样时刻的实际输出值y(k+1)；
(4)更新改控制器中的隶属度函数中心c、基宽b和网络权值w；
(5)令k＝k+1，移到下一采样时刻，返回步骤(1)，再进行重新计算。
内环控制回路控制方法步骤如下：
步骤B1：通过副变送器对减温器进行温度采集，获取减温器温度信号；
步骤B2：副控制器根据外环回路控制信号、温度反馈补偿信号以及减温器温度信号产生内环控制信号；
步骤B3：执行器根据内环控制信号动作，完成内环控制回路控制。
步骤A3中还包括：论域调整器根据温度偏差量、温度偏差变化率，调节模糊RBF神经网络中的伸缩因子；
神经网络参数调整器调节模糊RBF神经网络中的连接权、隶属度函数中心和基宽。
模糊RBF神经网络是由神经网络结构实现的模糊控制算法，利用神经网络 参数调整器可以确定和改变模糊RBF神经网络的连接权wij、隶属度函数中心cij和基宽bij，该控制器的输出为：
Δu(k)＝k pe(k)+k i[e(k)-e(k-1)]+k d[e(k)-2e(k-1)+e(k-2)]
选择增量式PID算法为：
u(k)＝u(k-1)+Δu(k)
其中e(k)为第k次采样时刻的系统偏差，u(k)为第k次采样时刻的输出值，Δu(k)为第k次采样时刻的输出增量；
该系统采用的是有监督学习算法，定义学习的目标函数为：
Figure PCTCN2022120450-appb-000001
其中r(k)和y(k)分别为该系统在kT时刻的理想输出与实际输出，r(k)-y(k)表示为迭代步骤k的控制误差；
上述控制方法采用梯度下降法进行搜索寻优，输出权wij、隶属度函数中心cij和基宽bij的迭代算法为：
Figure PCTCN2022120450-appb-000002
Figure PCTCN2022120450-appb-000003
Figure PCTCN2022120450-appb-000004
其中，k为迭代步骤；α为惯性系数，α∈[0,1]；η为学习速率，η∈[0,1]。
步骤B3中执行器根据内环控制信号动作具体为：执行器对设置在执行器与减温器相连通的减温水管道上的减温阀进行调节，调节流入减温水管道中的减温水的流量。
虽然结合附图描述了本发明的实施例，但是本领域技术人员可以在不脱离 本发明的精神和范围的情况下作出各种修改和变型，这样的修改和变型均落入由所附权利要求所限定的范围之内。

Claims (9)

  1. 一种基于事件触发的模糊神经网络温度控制系统，其特征在于，包括：外环控制回路、内环控制回路，所述外环控制回路与内环控制回路构成串级控制回路；
    所述外环控制回路包括：主控制器、主蒸汽对象、主变送器；
    所述主控制器包括：事件触发器、模糊RBF神经网络、PID控制器；
    所述内环控制回路包括：副控制器、执行器、减温器、副变送器；
    所述主蒸汽对象的输出端与所述主变送器的输入端连接；所述主变送器的输出端经与所述事件触发器的输入端连接；所述事件触发器的输出端与模糊RBF神经网络的输入端连接；所述模糊RBF神经网络输出端与所述PID控制器的输入端连接；所述PID控制器的输出端与所述副控制器的输入端连接；所述副控制器的输出端与所述执行器的输入端连接；所述执行器的输出端与所述减温器的输入端连接；所述减温器的输出端分别与所述主蒸汽对象的输入端、副变送器的输入端连接；所述副变送器的输出端与所述副控制器的输入端连接。
  2. 如权利要求1所述的基于事件触发的模糊神经网络温度控制系统，其特征在于，所述外环控制回路还包括：状态观测器；
    所述主变送器的输出端与所述状态观测器的输入端连接；所述状态观测器的输出端分别与所述事件触发器的输入端、副控制器的输入端连接。
  3. 如权利要求1所述的基于事件触发的模糊神经网络温度控制系统，其特征在于，所述主控制器还包括：论域调整器、神经网络参数调整器；
    所述状态观测器的输出端与所述论域调整器的输入端连接；所述论域调整器的输出端与所述模糊RBF神经网络的输入端连接；所述神经网络参数调整器 与所述模糊RBF神经网络连接。
  4. 一种基于事件触发的模糊神经网络温度控制系统的控制方法，其特征在于，包括：外环控制回路控制方法、内环控制回路控制方法；其中：
    所述外环控制回路控制方法步骤如下：
    步骤A1：通过主变送器对主蒸汽对象进行温度采集，获取主蒸汽温度信号；
    步骤A2：将主蒸汽温度信号与标准温度信号进行比较，计算温度偏差量、温度偏差变化率；
    步骤A3：事件触发器根据温度偏差变化率进行事件触发判断，当事件触发时，事件触发器输出接收的温度偏差变化率；当事件不触发时，事件触发器不输出信号；
    步骤A4：当模糊RBF神经网络接收到温度偏差变化率时，根据模糊RBF神经网络规则对PID控制器的三个参数进行整定，直至PID控制器的参数达到最优；
    步骤A5：PID控制器根据控制参数输出外环回路控制信号，完成外环控制回路控制；
    所述内环控制回路控制方法步骤如下：
    步骤B1：通过副变送器对减温器进行温度采集，获取减温器温度信号；
    步骤B2：副控制器根据所述外环回路控制信号以及减温器温度信号产生内环控制信号；
    步骤B3：执行器根据内环控制信号动作，完成内环控制回路控制。
  5. 如权利要求4基于事件触发的模糊神经网络温度控制系统的控制方法，其特征在于，所述步骤A2中还包括：状态观测器根据主蒸汽温度信号产生温度反 馈补偿信号；
    所述步骤B2：副控制器根据所述外环回路控制信号、温度反馈补偿信号以及减温器温度信号产生内环控制信号。
  6. 如权利要求4基于事件触发的模糊神经网络温度控制系统的控制方法，其特征在于，所述步骤A3中还包括：论域调整器根据温度偏差量、温度偏差变化率，调节模糊RBF神经网络中的伸缩因子；
    神经网络参数调整器调节模糊RBF神经网络中的连接权、隶属度函数中心和基宽。
  7. 如权利要求4基于事件触发的模糊神经网络温度控制系统的控制方法，其特征在于，所述事件触发器的触发条件为：
    ||de/dt((k+i)h)-de/dt(kh)||≤σ
    其中，de/dt((k+i)h)是当前时刻(k+i)的温度偏差变化率，de/dt(kh)是上一时刻(k)的温度偏差变化率，||||表示范数，σ为(0,1)的有界正数，i＝1,2,…,为正整数。
  8. 如权利要求4基于事件触发的模糊神经网络温度控制系统的控制方法，其特征在于，所述步骤B3中执行器根据内环控制信号动作具体为：执行器对设置在执行器与减温器相连通的减温水管道上的减温阀进行调节，调节流入减温水管道中的减温水的流量。
  9. 如权利要求4基于事件触发的模糊神经网络温度控制系统的控制方法，其特征在于，所述副控制器为PI控制方式。
PCT/CN2022/120450 2021-11-22 2022-09-22 一种基于事件触发的模糊神经网络温度控制系统及方法 WO2023087906A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202111388986.0 2021-11-22
CN202111388986.0A CN114089795B (zh) 2021-11-22 2021-11-22 一种基于事件触发的模糊神经网络温度控制系统及方法

Publications (1)

Publication Number Publication Date
WO2023087906A1 true WO2023087906A1 (zh) 2023-05-25

Family

ID=80302996

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2022/120450 WO2023087906A1 (zh) 2021-11-22 2022-09-22 一种基于事件触发的模糊神经网络温度控制系统及方法

Country Status (2)

Country Link
CN (1) CN114089795B (zh)
WO (1) WO2023087906A1 (zh)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116594446A (zh) * 2023-07-19 2023-08-15 广州豪特节能环保科技股份有限公司 一种大数据中心温控方法及系统
CN117193430A (zh) * 2023-10-11 2023-12-08 深圳市顾美科技有限公司 基于plc的pid温度控制方法、设备及存储介质
CN117316356A (zh) * 2023-10-24 2023-12-29 中国民航大学 一种复合材料构件热压罐成型工艺参数前馈补偿调控方法

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114089795B (zh) * 2021-11-22 2022-08-16 江苏科技大学 一种基于事件触发的模糊神经网络温度控制系统及方法
CN115437425A (zh) * 2022-09-28 2022-12-06 深圳市汇川技术股份有限公司 温度控制方法、装置、设备以及存储介质

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105334888A (zh) * 2015-11-16 2016-02-17 江苏科技大学 一种基于触发机制的网络串级温度控制系统及其方法
CN109459928A (zh) * 2018-06-12 2019-03-12 陕西科技大学 模糊分数阶PIDμ控制器的DDS置换蒸煮温度控制方法
CN111812968A (zh) * 2020-06-24 2020-10-23 合肥工业大学 基于模糊神经网络pid控制器的阀位串级控制方法
CN112799297A (zh) * 2020-11-11 2021-05-14 华能国际电力股份有限公司营口电厂 一种温度预测控制方法、系统、设备及可读存储介质
CN114089795A (zh) * 2021-11-22 2022-02-25 江苏科技大学 一种基于事件触发的模糊神经网络温度控制系统及方法

Family Cites Families (17)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7117045B2 (en) * 2001-09-08 2006-10-03 Colorado State University Research Foundation Combined proportional plus integral (PI) and neural network (nN) controller
GB0710524D0 (en) * 2007-06-01 2007-07-11 Lysanda Ltd Improvements in and relating to engine monitoring
CN101763035A (zh) * 2009-11-13 2010-06-30 上海电力学院 Rbf神经网络整定pid与模糊免疫控制方法
JP6222234B2 (ja) * 2013-09-30 2017-11-01 理化工業株式会社 制御装置及び制御方法
JP6269678B2 (ja) * 2013-09-30 2018-01-31 理化工業株式会社 ファジイ制御装置及びファジイ制御方法
CN104201955A (zh) * 2014-08-12 2014-12-10 中国南方电网有限责任公司超高压输电公司广州局 一种基于变论域模糊的特高压换流变分接开关控制方法
CN105546508B (zh) * 2016-02-18 2017-10-31 江苏科技大学 基于事件触发机制的火电厂主蒸汽温度控制系统及方法
CN107491113A (zh) * 2017-04-19 2017-12-19 安徽华脉科技发展有限公司 一种基于arm的模糊pid温度控制系统
CN107390528A (zh) * 2017-08-23 2017-11-24 华南理工大学 一种焊缝跟踪应用的自适应模糊控制方法
CN108303888B (zh) * 2018-02-07 2020-11-03 广东电网有限责任公司电力科学研究院 一种电站锅炉主蒸汽温度减温喷水控制方法及系统
CN109120198A (zh) * 2018-09-11 2019-01-01 江苏科技大学 基于触发机制的电机控制系统及方法
CN109143872A (zh) * 2018-11-19 2019-01-04 重庆科技学院 一种基于事件触发gdhp的连续搅拌反应釜过程控制方法
CN109733242B (zh) * 2018-12-12 2022-05-10 西北工业大学 电动汽车充电机的神经模糊稳定性控制系统及控制方法
CN110554715A (zh) * 2019-10-25 2019-12-10 攀钢集团攀枝花钢铁研究院有限公司 基于rbf神经网络的硫酸氧钛外加晶种水解工艺温度pid控制方法
CN111983918A (zh) * 2020-09-01 2020-11-24 南通大学 一种基于改进型模糊Smith-PID的电加热炉温度控制方法
CN112286051A (zh) * 2020-09-20 2021-01-29 国网江苏省电力有限公司信息通信分公司 复杂网络攻击下基于自适应事件触发机制的神经网络量化控制方法
CN113211446B (zh) * 2021-05-20 2023-12-08 长春工业大学 一种事件触发-神经动态规划的机械臂分散跟踪控制方法

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105334888A (zh) * 2015-11-16 2016-02-17 江苏科技大学 一种基于触发机制的网络串级温度控制系统及其方法
CN109459928A (zh) * 2018-06-12 2019-03-12 陕西科技大学 模糊分数阶PIDμ控制器的DDS置换蒸煮温度控制方法
CN111812968A (zh) * 2020-06-24 2020-10-23 合肥工业大学 基于模糊神经网络pid控制器的阀位串级控制方法
CN112799297A (zh) * 2020-11-11 2021-05-14 华能国际电力股份有限公司营口电厂 一种温度预测控制方法、系统、设备及可读存储介质
CN114089795A (zh) * 2021-11-22 2022-02-25 江苏科技大学 一种基于事件触发的模糊神经网络温度控制系统及方法

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
"Master's Thesis", 3 April 2012, HENAN POLYTECHNIC UNIVERSITY, CN, article LIU, LULU: "Research of Control Policy on Main Steam Temperature of Thermal Power Plant Based on Fuzzy Neural Network", pages: 1 - 67, XP009545877 *

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116594446A (zh) * 2023-07-19 2023-08-15 广州豪特节能环保科技股份有限公司 一种大数据中心温控方法及系统
CN116594446B (zh) * 2023-07-19 2023-10-17 广州豪特节能环保科技股份有限公司 一种大数据中心温控方法及系统
CN117193430A (zh) * 2023-10-11 2023-12-08 深圳市顾美科技有限公司 基于plc的pid温度控制方法、设备及存储介质
CN117193430B (zh) * 2023-10-11 2024-05-24 深圳市顾美科技有限公司 基于plc的pid温度控制方法、设备及存储介质
CN117316356A (zh) * 2023-10-24 2023-12-29 中国民航大学 一种复合材料构件热压罐成型工艺参数前馈补偿调控方法
CN117316356B (zh) * 2023-10-24 2024-05-17 中国民航大学 一种复合材料构件热压罐成型工艺参数前馈补偿调控方法

Also Published As

Publication number Publication date
CN114089795B (zh) 2022-08-16
CN114089795A (zh) 2022-02-25

Similar Documents

Publication Publication Date Title
WO2023087906A1 (zh) 一种基于事件触发的模糊神经网络温度控制系统及方法
CN104482525B (zh) 超超临界机组再热汽温的控制方法和系统
CN107218594B (zh) 锅炉主蒸汽温度多参量智能控制系统
CN107178778B (zh) 锅炉汽温控制装置、系统和方法
CN106919053A (zh) 一种基于变结构预测控制算法的火电机组协调控制系统
Liu et al. Intelligent decoupling control of power plant main steam pressure and power output
CN108361683B (zh) 一种全负荷段再热气温智能控制系统
CN103134046B (zh) 一种火电机组过热汽温两级协调预测控制方法
CN111765447B (zh) 一种基于多变量解耦的发电锅炉主汽温控制方法和系统
CN105546508A (zh) 基于事件触发机制的火电厂主蒸汽温度控制系统及方法
CN106933202A (zh) 利用基于所估计的状态信息的间歇重新初始化的前馈控制
CN113359425A (zh) 一种基于lstm神经网络pid优化的火电厂锅炉主汽温智能控制系统
CN111102559B (zh) 一种基于双神经网络逆模型的屛式过热器汽温控制方法
Alamoodi et al. Nonlinear control of coal-fired steam power plants
CN113835342A (zh) 一种过热汽温系统的抗扰预测控制方法
Sun et al. DEB-oriented modelling and control of coal-fired power plant
CN113883492B (zh) 锅炉蒸汽汽温控制方法及电子设备
CN110631002A (zh) 一种火电机组主气温的控制方法
Han et al. A L1-LEMPC hierarchical control structure for economic load-tracking of super-critical power plants
Hu et al. Feedforward DMC-PID cascade strategy for main steam temperature control system in fossil-fired power plant
Ma et al. Intelligent Compensation for the Set Values of PID Controllers to Improve Boiler Superheated Steam Temperature Control
Jing et al. Application of improved model-free adaptive control in an industrial oiler system
CN216281315U (zh) 一种双渣室燃煤机组主蒸汽温度优化控制装置
CN117250866A (zh) 一种pgnn前馈动态补偿-多模型预测控制方法
CN114115376B (zh) 基于事件触发的神经网络预测串级温度控制系统及其方法

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 22894444

Country of ref document: EP

Kind code of ref document: A1