Torch combustion control method, device, equipment and medium

Abstract

The invention relates to the technical field of torch combustion. The invention discloses a torch combustion control method, device and equipment and a medium. The method comprises the steps that the outlet pressure of a booster pump, the temperature of a fuel pipeline and the front end pressure of a burner nozzle are obtained; the booster pump outlet pressure, the fuel pipeline temperature and the burner nozzle front end pressure serve as input values to be input into a pre-trained neural network model, and output values of the pre-trained neural network model are three adjustable parameters Kp, Ki and Kd of a PID controller; wherein when the neural network model is trained, a weight and a threshold value of the neural network model are updated by using a fox-monkey algorithm; and the frequency of the booster pump is adjusted according to the three adjustable parameters Kp, Ki and Kd of the PID controller. Timely adjustment of the frequency of the booster pump is achieved, automatic control over torch combustion is improved, and energy waste is reduced.

Description

Technical Field

[0001] The present application relates to the technical field of flare combustion, and in particular to a flare combustion control method, device, equipment and medium. Background Art

[0002] The main torch is the focus of public attention at the opening ceremony of various sports games. The flare combustion control system, as an important component of the main torch, plays a vital role in the safety of the main torch operation and guarantee, the accuracy of torch ignition, the reliability of combustion, the stability of operation under extreme weather conditions, the innovation and leadership of the concept, etc. However, the adjustment strategy of the existing flare combustion system is still limited to manual adjustment, with a low degree of automation, resulting in energy waste. Summary of the invention

[0003] In view of the above situation, the embodiments of the present application provide a flare combustion control method, device, equipment and medium, which aim to solve the above problems or at least partially solve the above problems.

[0004] In a first aspect, the present application provides a flare combustion control method, comprising:

[0005] Obtain booster pump outlet pressure, fuel pipeline temperature, and burner nozzle front end pressure;

[0006] The boost pump outlet pressure, the fuel pipeline temperature, and the burner nozzle front end pressure are input as input values ​​into a pre-trained neural network model, and the output values ​​of the pre-trained neural network model are three adjustable parameters Kp, Ki, and Kd of a PID controller;

[0007] The frequency of the boost pump is adjusted according to the three adjustable parameters Kp, Ki, and Kd of the PID controller.

[0008] Preferably, the method comprises:

[0009] When training the neural network model, the weights and thresholds of the neural network model are updated using the Lemur algorithm, and the updating process includes:

[0010] Get Lemur algorithm parameters;

[0011] The matrix consisting of booster pump outlet pressure, fuel pipeline temperature, burner nozzle front end pressure, weight initial value and threshold initial value is used as the lemur population set;

[0012] Determine an initial position for starting iteration, and calculate an objective function for each lemur in the lemur population set in each iteration, wherein the objective function is the performance of the neural network model under current weights and thresholds;

[0013] According to the Lemur algorithm parameters, the free risk rate is calculated in each iteration;

[0014] According to the objective function of each lemur and the free risk rate in each iteration, updating the global optimal lemur and the nearest lemur of each lemur;

[0015] Obtain a random number, and update the lemur population set according to the random number and the free risk rate;

[0016] Based on the preset output conditions, the weights and thresholds represented by the global best lemur are output.

[0017] Preferably, the outputting of the weight and threshold represented by the global best lemur based on the preset output condition comprises:

[0018] The jump rate is obtained, and when the random number is greater than or equal to the jump rate, the weight and threshold represented by the global best lemur are output.

[0019] Preferably, the method further comprises:

[0020] In the nth cycle, the return liquid in the burner is controlled to flow back to the first buffer tank through the return liquid pipeline, and the second buffer tank is replenished with liquid, and the fuel in the second buffer tank is sent to the burner through the booster pump and the liquid supply pipeline;

[0021] In the (n+1)th cycle, the return liquid in the burner is controlled to flow back to the second buffer tank through the return liquid pipeline, and the first buffer tank is replenished with liquid, and the fuel in the first buffer tank is sent to the burner through the booster pump and the liquid supply pipeline.

[0022] Preferably, the method further comprises:

[0023] Obtain the first temperature rise data of the fuel pipeline when the boost pump has different frequencies and the return valve has the same opening;

[0024] Obtain the second temperature rise data of the fuel pipeline when the boost pump has the same frequency and the return valve has different openings;

[0025] Obtaining a functional relationship between a boost pump frequency and a return valve opening according to the first temperature rise data of the fuel pipeline and the second temperature rise data of the fuel pipeline;

[0026] The frequency of the boost pump and the opening of the return valve are adjusted according to the functional relationship between the frequency of the boost pump and the opening of the return valve.

[0027] Preferably, the method further comprises:

[0028] Acquire liquid level information in the first buffer tank and the second buffer tank;

[0029] Controlling a first liquid replenishing pump to replenish liquid into the first buffer tank according to the liquid level information of the first buffer tank;

[0030] According to the liquid level information of the second buffer tank, the second liquid replenishing pump is controlled to replenish liquid into the second buffer tank.

[0031] Preferably, the method further comprises:

[0032] Obtain the pressure difference data and fuel concentration data at the door of the explosion-proof room;

[0033] The explosion-proof fan is controlled to start, change speed or stop according to the pressure difference data at the explosion-proof room door and the fuel concentration data.

[0034] In a second aspect, a flare combustion control device is provided, comprising:

[0035] An acquisition module is used to obtain the boost pump outlet pressure, fuel pipeline temperature, and burner nozzle front end pressure;

[0036] A neural network module, used for inputting the boost pump outlet pressure, the fuel pipeline temperature, and the burner nozzle front end pressure as input values ​​into a pre-trained neural network model, wherein the output values ​​of the pre-trained neural network model are three adjustable parameters Kp, Ki, and Kd of the PID controller; wherein the weights and thresholds of the neural network model are updated using the Lemur algorithm when training the neural network model;

[0037] The adjustment module is used to adjust the boost pump frequency according to the three adjustable parameters Kp, Ki, and Kd of the PID controller.

[0038] In a third aspect, a computer device is provided, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the flare combustion control method as described in the first aspect when executing the computer program.

[0039] In a fourth aspect, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the flare combustion control method as described in the first aspect are implemented.

[0040] At least one of the above technical solutions adopted in the embodiments of the present application can achieve the following beneficial effects:

[0041] The present application obtains the boost pump outlet pressure, fuel pipeline temperature, and burner nozzle front end pressure into a pre-trained neural network model to obtain the output value of the pre-trained neural network model: three adjustable parameters Kp, Ki, and Kd of the PID controller. According to the output value, timely adjustment of the boost pump frequency is achieved, thereby improving the automated control of flare combustion and reducing energy waste. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0043] Figure 1 is a schematic diagram of a flare combustion control system in one embodiment of the present invention;

[0044] Figure 2 is a functional diagram of a PLC module group in one embodiment of the present invention;

[0045] Figure 3 is an electrical diagram of a PLC module group in one embodiment of the present invention;

[0046] Figure 4 is a schematic diagram of an application environment of a flare combustion control method in one embodiment of the present invention;

[0047] Figure 5 It is a schematic flow chart of a flare combustion control method in one embodiment of the present invention;

[0048] Figure 6 It is a schematic diagram of a process of updating the weights and thresholds of a neural network model using a lemur algorithm in one embodiment of the present invention;

[0049] Figure 7 is a structural schematic diagram of a flare combustion control device in one embodiment of the present invention;

[0050] Figure 8 is a schematic diagram of a structure of a computer device in one embodiment of the present invention;

[0051] Fig. 9 It is another structural schematic diagram of a computer device in one embodiment of the present invention. DETAILED DESCRIPTION

[0052] In order to make the purpose, technical solution and advantages of the present application clearer, the technical solution of the present application will be clearly and completely described below in combination with the specific embodiments of the present application and the corresponding drawings. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present application.

[0053] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that such use is interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the term "including" and its variants are to be interpreted as open-ended terms meaning "including but not limited to".

[0054] The technical solutions provided by various embodiments of the present application are described in detail below in conjunction with the accompanying drawings.

[0055] As mentioned above, the current adjustment strategy of the flare combustion system is still limited to manual adjustment, with a low degree of automation, resulting in energy waste. In order to solve this technical problem, the present application embodiment provides a flare combustion control method, which is applied in Figure 1 The flare combustion system shown.

[0056] Embodiment 1

[0057] like Figure 1 As shown, the flare combustion system includes a MCGS configuration monitoring system 100, a control cabinet system 200, a fuel supply device 300, a fuel delivery device 500, and a combustion device 400.

[0058] The control cabinet system 200 includes a frequency conversion cabinet 210 and a PLC module group 220 , and the frequency conversion cabinet 210 and the PLC module group 220 are electrically connected to the MCGS configuration monitoring system 100 respectively.

[0059] The fuel supply device 300 includes a supply control device 330, a liquid storage tank 310 and a liquid replenishing pump, the liquid replenishing pump includes a first liquid replenishing pump 321 and a second liquid replenishing pump 322, the supply control device 330 and the PLC module group 220 are electrically connected, the liquid storage tank 310 is used to store fuel, and a liquid level sensor, a temperature sensor and a pressure sensor are arranged in the liquid storage tank 310, and the liquid level sensor, the temperature sensor and the pressure sensor in the liquid storage tank 310 are electrically connected to the supply control device 330 respectively, the liquid replenishing pump is arranged in the liquid storage tank 310, and the liquid replenishing pump is electrically connected to the supply control device 330.

[0060] The fuel delivery device 500 includes a buffer tank, a booster pump, a fuel pipeline and an explosion-proof fan group 530. The buffer tank includes a first buffer tank 511 and a second buffer tank 512. A plurality of booster pumps are provided, each of which is provided on a fuel pipeline to deliver fuel to different combustion heads. The booster pumps include a first booster pump 521, a second booster pump 522, a third booster pump 523, a fourth booster pump 524, a fifth booster pump 525 and a sixth booster pump 526. The first booster pump 521, the second booster pump 522, the third booster pump 523, the fourth booster pump 524, the fifth booster pump 525 and the sixth booster pump 526 are connected to the inlet and outlet of the two buffer tanks respectively and are arranged in parallel. The buffer tank and the liquid storage tank 310 are connected through a liquid replenishing pump, and the fuel in the liquid storage tank 310 is delivered to the buffer tank through the liquid replenishing pump, and the fuel in the buffer tank is delivered to the combustion device 400 through the booster pump and the fuel pipeline. The fuel delivery device 500 also includes a liquid level sensor, a temperature sensor and a pressure sensor arranged in the buffer tank, a pressure sensor, a temperature sensor and a frequency converter arranged on the boosting pump, the frequency converter and the frequency converter cabinet 210 are electrically connected, and the liquid level sensor, temperature sensor and pressure sensor arranged in the buffer tank, the explosion-proof fan group 530, the pressure sensor and the temperature sensor arranged on the boosting pump are electrically connected to the PLC module group 220. The boosting pump uses a stainless steel magnetic vortex pump, which is corrosion-resistant, has a small flow rate and a high head. The frequency converter is used to modulate the frequency of the boosting pump to ensure the constant pressure of the fuel pipeline and suppress the temperature rise of the buffer tank. The explosion-proof fan group 530 is provided with a supply fan and an exhaust fan, and the supply fan, the exhaust fan and the PLC module group 220 are electrically connected. The interlocking control is set according to the concentration of the fuel detector, and the exhaust fan can switch between high speed and low speed according to the concentration of the fuel detector.

[0061] The combustion device 400 includes a combustion head, a combustion-supporting fan and a camera. The combustion head includes a first group of combustion heads 411, a second group of combustion heads 412, a third group of combustion heads 413, and a fourth group of combustion heads 414. The first group of combustion heads 411, the second group of combustion heads 412, the third group of combustion heads 413, and the fourth group of combustion heads 414 constitute four circles of combustion heads of the torch, see Figure 1As shown. The fuel pipeline is connected to the combustion head, and each group of the combustion head is composed of an atomizing nozzle, a solenoid valve, an igniter, and a pressure sensor. The combustion-supporting fan 430, the camera 420, the solenoid valve in the combustion head, the igniter, the pressure sensor and the PLC module group 220 are electrically connected. The camera 420 monitors the operation status of the combustion device 400 in real time, including the flame, the solenoid valve, the pipeline and whether there is fuel leakage. In order to prevent the camera 420 from blacking out the image due to the high voltage interference of the igniter at the moment of ignition, the power supply of the camera 420 is equipped with a power filter. During the torch combustion process, the number of combustion-supporting fans 430 is controlled as needed to enhance the flame combustion effect and increase the wind and rain resistance. The solenoid valve of the combustion head adopts a direct-acting structure, and the use pressure is zero pressure difference, which solves the problem of leakage of the atomizing nozzle during non-use when the solenoid valve is installed vertically or inverted. The pressure sensor of the combustion head detects the pressure at the front end of the burner nozzle in real time to ensure the fuel atomization effect of the atomizing nozzle, effectively control the amount of unburned fuel, reduce energy consumption, and reduce safety hazards.

[0062] The principle of the torch combustion system is: the fuel in the liquid storage tank 310 is sent to the buffer tank through the liquid replenishing pump, and the fuel in the buffer tank is sent to the combustion head through the fuel pipeline by the booster pump, triggering the igniter, the fuel at the combustion head is ignited, and the torch is ignited successfully.

[0063] One liquid storage tank 310 is provided, two liquid replenishing pumps are provided, and two buffer tanks are provided, and the liquid replenishing pumps and buffer tanks are provided one by one. Multiple booster pumps are provided, and the multiple booster pumps are connected to the two buffer tanks respectively. Each booster pump is connected to multiple burner heads correspondingly, and by controlling the number of booster pumps turned on, the number of burner heads ignited can be controlled, and the size of the flame can be adjusted.

[0064] The MCGS configuration monitoring system 100 includes a monitoring interface, which displays in real time the operating parameters of the flare combustion system collected by the data acquisition PLC module, including the liquid level, temperature and pressure of the buffer tank, the temperature of the fuel pipeline, the outlet pressure of the booster pump, the front end pressure of the burner nozzle, the concentration of the fuel detector, the pressure difference at the door of the explosion-proof room, etc. The MCGS configuration monitoring system 100 also includes a setting interface, which sets the parameters of the flare combustion system, including the ignition time setting of the igniter, the high, medium and low limits of the buffer tank liquid level, the temperature and pressure upper limits, the fuel pipeline temperature upper limit, the booster pump outlet pressure upper limit, the fuel detector concentration upper limit, the pressure difference between the two doors of the explosion-proof room, the target pressure value of the burner fuel atomization, the system automatic control or manual control selection, and the start and selection of the replenishment pump.

[0065] The MCGS configuration monitoring system 100 is set on the industrial computer 110. The MCGS configuration monitoring system 100 can realize manual testing and starting and stopping of each booster pump, solenoid valve, combustion-supporting fan 430, explosion-proof fan group 530, and igniter in the flare combustion system.

[0066] The MCGS configuration monitoring system 100 can complete the pressure building of the flare combustion system 2 minutes before the flare is ignited through the monitoring interface. Start the booster pump, and when the pressure at the front end of the burner nozzle meets the preset atomization working pressure, the flare combustion system is ignited through the monitoring interface. During the ignition period, the buttons on the monitoring interface can be used to switch between low fire, high fire, and flameout. This process can be completed by controlling the start and stop times of the booster pump and the solenoid valve.

[0067] The MCGS configuration monitoring system 100 also includes an alarm interface that can record in real time the fault alarm status of equipment in the flare combustion system, including booster pump failure, fire detection failure, explosion-proof fan group 530 failure, and sensor readings exceeding the limit.

[0068] During operation, the MCGS configuration monitoring system 100 stores the operating parameters and sensor readings in real time, generates real-time curves, and understands the working status of the flare combustion system by comparing the real-time curves with the historical curves, and analyzes the combustion trend of the flare combustion system. It can also view and export key data in daily reports, and the types of key data can be set according to user needs.

[0069] The control cabinet system 200 is provided with a PLC module group 220 and a frequency conversion cabinet 210. Figure 2 and Figure 3 As shown, the PLC module group 220 is composed of a control PLC module and a collection PLC module group 220. In this example, the control PLC module uses a Siemens SMART series 1 SR20 host, 3 QR16-IO expansion modules, and 2 AQ04 analog output modules, and the collection PLC module uses a Siemens SMART series 1 SR20 host and 6 AE04 analog input modules.

[0070] The control PLC module 211 receives the operation signal and fault signal of the booster pump through the SR20 host, and receives the control signals of each igniter of the combustion head. The QR16-IO expansion module of the control PLC module receives the start and stop control signals of each solenoid valve of the combustion head, the replenishment pump, and the booster pump control signal. The AQ04 analog output module of the control PLC module receives the target frequency setting of the inverter in the frequency conversion cabinet 210 corresponding to each booster pump.

[0071] The acquisition PLC module is connected to the operation signal and fault signal of the explosion-proof fan group 530 through the SR20 host, and receives the control signal of the combustion-supporting fan 430, the fuel concentration over-limit alarm signal, and the explosion-proof fan group 530 control signal. The acquisition PLC module's AE04 analog input module is connected to the liquid level, temperature, and pressure signals of the buffer tank, the fuel pipeline temperature signal, the boost pump outlet pressure signal, the return pipe pressure signal, the burner nozzle front end pressure signal, the explosion-proof room fuel concentration signal, and the explosion-proof room door pressure difference signal.

[0072] During the ignition process of the torch combustion system, according to weather conditions and the requirements of the flame combustion posture, the combustion-supporting fan 430 can be turned on by controlling the PLC module to make the fuel atomization and combustion more complete, so that the burning flame meets the viewing needs while also increasing the ability to resist wind and rain.

[0073] The sensor values ​​collected in real time by the PLC module are transmitted to the MCGS configuration monitoring system 100 via the network. The monitoring interface of the MCGS configuration monitoring system 100 displays the current value of each sensor. The MCGS configuration monitoring system 100 can save the collected sensor data in real time, query reports, and draw real-time and historical curves to reflect the working status and operation trend of the flare combustion system.

[0074] The acquisition PLC module realizes data sharing with the supply control device 330 through Modbus communication. The MCGS configuration software system monitoring interface can display the liquid level and pressure of the liquid storage tank 310 and the operating status (start, stop, frequency conversion) of the replenishment pump in real time. The acquisition PLC module transmits the liquid level and pressure of the buffer tank and the concentration parameters of the fuel detector to the supply control device 330 in real time, so that the supply control device 330 can adjust the operating status of the replenishment pump in time. In order to solve the false alarm of fuel concentration caused by power failure in the explosion-proof room during unattended use other than drills, the acquisition PLC module adds a watchdog function.

[0075] There are two sets of control PLC modules and acquisition PLC modules, which serve as backup redundant control for each other.

[0076] In order to improve the environmental performance of the flare combustion system, methanol is selected as the fuel in the flare combustion system in this embodiment.

[0077] Embodiment 2

[0078] The flare combustion control method provided by the embodiment of the present invention can be applied in Figure 4In the application environment, the device side communicates with the server side through the network. The server side can obtain the boost pump outlet pressure, the fuel pipeline temperature, and the burner nozzle front end pressure through the device side; the boost pump outlet pressure, the fuel pipeline temperature, and the burner nozzle front end pressure are input into the pre-trained neural network model as input values, and the output value of the pre-trained neural network model is the three adjustable parameters Kp, Ki, and Kd of the PID controller; wherein, when training the neural network model, the lemur algorithm is used to update the weights and thresholds of the neural network model; the boost pump frequency is adjusted according to the three adjustable parameters Kp, Ki, and Kd of the PID controller. Among them, the device side is the data collected by various sensors in the torch combustion system. The server can be implemented with an independent server or a server cluster composed of multiple servers. In this embodiment, the server can also be an MCGS configuration software system.

[0079] The present invention is described in detail below through specific embodiments.

[0080] See also Figure 5 As shown, Figure 5 A schematic flow chart of a flare combustion control method provided by an embodiment of the present invention comprises the following steps:

[0081] S10: Obtaining the boost pump outlet pressure, fuel pipeline temperature, and burner nozzle front end pressure;

[0082] S20: inputting the boost pump outlet pressure, the fuel pipeline temperature, and the burner nozzle front end pressure as input values ​​into a pre-trained neural network model, wherein the output values ​​of the pre-trained neural network model are three adjustable parameters Kp, Ki, and Kd of the PID controller; wherein, when training the neural network model, the weights and thresholds of the neural network model are updated using the Lemur algorithm;

[0083] S30: adjusting the frequency of the boost pump according to the three adjustable parameters Kp, Ki, Kd of the PID controller.

[0084] Specifically, the neural network model in the present application adopts a three-layer BP neural network model, including an input layer, a hidden layer and an output layer. The pre-trained neural network model is obtained by inputting the historical combustion data of the flare combustion system into the BP neural network model for training. During the training process, the system state variables corresponding to the three input layer nodes of the BP neural network model are: booster pump outlet pressure, burner nozzle front end pressure and fuel pipeline temperature, and the output layer nodes correspond to the three adjustable parameters Kp, Ki, Kd of the PID controller. Each time, an error analysis is performed based on the results obtained from the training and the results in the historical combustion data, and then the weights and thresholds are adjusted to gradually make the results obtained from the training consistent with the results in the historical combustion data.

[0085] Regarding the BP neural network model, the specific algorithm is as follows:

[0086] The input and output of the hidden layer of the BP neural network NN are

[0087]

[0088] in, is the hidden layer weight; f[] is the activation function, which takes the positive and negative symmetric Sigmoid function, f[x] = tanh(x).

[0089] The input and output of the output layer are:

[0090]

[0091] in, is the output layer weight; g[] is the activation function, which takes the positive Sigmoid function, g[x] = (1 / 2)(1+tanh(x)).

[0092] The performance index function is taken as:

[0093]

[0094] The output layer weight adjustment formula is:

[0095]

[0096] Where η is the learning rate, a is the inertia coefficient, and

[0097]

[0098] Similarly, the hidden layer weight adjustment formula can be obtained as:

[0099]

[0100] in,

[0101]

[0102] In the above formula, the superscripts (1), (2), and (3) represent three-layer networks respectively.

[0103] During training, according to the selected BP neural network structure, first give the initial value of each layer weight and threshold, select the learning rate and inertia coefficient, and then calculate the three adjustable parameters Kp, Ki, Kd of the PID controller of the output layer according to the above formula.

[0104] When applied, the boost pump outlet pressure, the burner nozzle front end pressure and the fuel pipeline temperature are input into the trained neural network model, and the three adjustable parameters Kp, Ki and Kd of the PID controller corresponding to the output layer node can be obtained. According to the three adjustable parameters Kp, Ki and Kd of the PID controller, the frequency of the frequency converter in the frequency converter cabinet 210 corresponding to each boost pump is controlled in a closed-loop manner in real time, thereby achieving the optimal control purpose of the system.

[0105] Regarding the Lemur algorithm, the introduction is as follows:

[0106] The Lemurs Optimizer (LO) algorithm is used to optimize the weights and thresholds of the BP neural network model, thereby improving the generalization ability and learning performance of the BP neural network, thereby improving the overall search efficiency of the BP neural network model and obtaining better PID adjustment parameters.

[0107] The Lemur Optimizer (LO) simulates the jumping and dancing behaviors of lemurs, and has the advantages of simple structure, novel ideas, and fast search speed. In this algorithm, two main behaviors of lemurs are used as inspiration: jumping and dancing, and two optimization stages of metaheuristics are used - search and exploitation, which are very similar to these two movement behaviors.

[0108] During the search phase, the Lemur algorithm uses the jumping behavior of lemurs to explore the search space and find the best position. Each solution is considered as a lemur, and each vector represents a separate coordinate of the lemur. At the same time, the best position is assigned to each solution, which is related to the fitness function value of the solution. Lemurs will change their position vectors based on the neatest lemur or the global optimal lemur. The main goal of the search phase is to let the lemurs jump to different areas to find the best position of the lemur in the search space. The set of lemurs is represented by a matrix, and its algorithm is a population-based algorithm, where T represents the set of lemurs in a population matrix of size s*d, d represents the decision variable, and s represents the candidate solution; the decision variables in the general solution m are randomly generated by the following formula:

[0109]

[0110] The function rand() generates a range of (1,2,...,MAX_INT). MAX_INT is the maximum integer that can be generated during the development phase. The discrete lower and upper limits of the variable n are [ub n ,lb n ]express;

[0111] During the development phase, lemurs with lower fitness values ​​tend to change their decision variables from lemurs with higher fitness values ​​to improve the overall fitness; in each iteration, lemurs are organized based on their fitness values ​​and the global optimal lemur (i.e., gbn ) and each lemur's closest lemur (i.e. bnl).

[0112] In this direction, each iteration uses the following two options to assign a value to the decision variable n in the solution m, namely, the value is selected from the global optimal lemur or the value is selected from the nearest lemur. The mathematical expression is:

[0113]

[0114] Among them, l(m,n) is the n value of the current lemur; l(bnl,n) is the n value of the best nearest lemur of the current lemur l(m,n); l(gbl,n) is the global optimal lemur, the free risk rate FRR is the risk rate of all lemurs in the group, and rand() represents a random number between [0,1];

[0115] Based on the above formula, it can be seen that the probability of FRR is the main coefficient of the Lemur algorithm, and the formula of this coefficient is:

[0116]

[0117] Among them, Low_Risk_Rate and High_Risk_Rate are fixed predefined values; MaxIter is the maximum number of iterations; CurrIter is the current number of iterations; It should be clear that the purpose of Low_Risk_Rate and High_Risk_Rate is to determine the minimum and maximum values ​​of FRR.

[0118] When training the neural network model, the weights and thresholds of the neural network model are updated using the Lemur algorithm, see Figure 6 include:

[0119] S01: Get the parameters of the Lemur algorithm; the parameters of the Lemur algorithm include the number of iterations MaxIter, the dimension (number of vector spaces), the number of solutions MAX_INT, the upper and lower limits [ub j ,lb j ] and low risk rate Low_Risk_Rate, high risk rate High_Risk_Rate.

[0120] S02: Using the matrix consisting of the booster pump outlet pressure training data, the fuel pipeline temperature training data, the burner nozzle front end pressure training data, the weight initial value and the threshold initial value as the lemur population set; specifically, combining Figure 1 As shown, it includes the booster pump outlet pressure value bP, bP1~bP6; the burner nozzle front end pressure value fP, fP1~fP4; the fuel pipeline temperature value pT, pT1~pT4; and also includes the constraint condition bP <bP s ,fP <fPs , pT <pT s (bP s ,fP s and pT s It is the set value of booster pump outlet pressure, burner nozzle front end pressure and fuel pipeline temperature).

[0121] S03: Determine the initial position for the start of iteration, and calculate the objective function of each lemur in the lemur population set in each iteration, wherein the objective function is the performance of the neural network model under the current weights and thresholds; specifically, the objective function is the performance of the neural network model under the current weights and thresholds, that is, the fitness value.

[0122] S04: Calculate the free risk rate in each iteration according to the Lemur algorithm parameters; specifically, calculate the free risk rate FRR according to formula (10).

[0123] S05: According to the objective function of each lemur and the free risk rate in each iteration, the global optimal lemur and the nearest lemur of each lemur are updated; specifically, the global optimal lemur and the nearest lemur of each lemur are updated according to formula (9).

[0124] S06: Obtain a random number, and update the lemur population set according to the random number and the free risk rate; specifically, obtain the random number rand according to formula (8). According to the FRR and the random number, decide whether to use the value of the global optimal lemur or the value of the nearest lemur to update the lemur population set.

[0125] S07: Based on the preset output conditions, the weight and threshold represented by the global best lemur are output.

[0126] Specifically, they include:

[0127] Obtain a jumping rate JumpingRate, and when the random number rand is greater than or equal to the jumping rate JumpingRate, output the weight and threshold represented by the global best lemur. When the random number rand is less than the jumping rate JumpingRate, repeat steps S04 to S07.

[0128] The output weights and thresholds are input into the BP neural network model being trained to optimize the BP neural network model.

[0129] Specifically, the training includes training data and test data, both of which are historical data, which are used for training and testing of the BP neural network model respectively. The training data includes training data of boost pump outlet pressure, fuel pipeline temperature, burner nozzle front end pressure, and training data of three adjustable parameters Kp, Ki, and Kd of the PID controller; the test data includes test data of boost pump outlet pressure, fuel pipeline temperature, burner nozzle front end pressure, and test data of three adjustable parameters Kp, Ki, and Kd of the PID controller.

[0130] Specifically, two buffer tanks are provided, including a first buffer tank 511 and a second buffer tank 512. Figure 1 The method also includes that in the low-fire mode, the buffer tank is put into use in a scheduling rotation manner to prevent the buffer tank, the booster pump and the fuel pipeline from rising too fast in temperature, and the method includes:

[0131] In the nth cycle, the return liquid in the burner is controlled to flow back to the first buffer tank 511 through the return liquid pipeline, and the second buffer tank 512 is replenished with liquid, and the fuel in the second buffer tank 512 is sent to the burner through the booster pump and the liquid supply pipeline;

[0132] In the n+1th cycle, the return liquid in the burner is controlled to flow back to the second buffer tank 512 through the return liquid pipeline, and the first buffer tank 511 is replenished with liquid, and the fuel in the first buffer tank 511 is sent to the burner through the booster pump and the liquid supply pipeline.

[0133] Specifically, when the buffer tank adopts a scheduling rotation method, the method further includes:

[0134] Obtain the first temperature rise data of the fuel pipeline when the boost pump has different frequencies and the return valve has the same opening;

[0135] Obtain the second temperature rise data of the fuel pipeline when the boost pump has the same frequency and the return valve has different openings;

[0136] Obtaining a functional relationship between a boost pump frequency and a return valve opening according to the first temperature rise data of the fuel pipeline and the second temperature rise data of the fuel pipeline;

[0137] The frequency of the boost pump and the opening of the return valve are adjusted according to the functional relationship between the frequency of the boost pump and the opening of the return valve.

[0138] Specifically, according to the first temperature rise data of the fuel pipeline and the second temperature rise data of the fuel pipeline, the temperature rise of the fuel pipeline is set within 45°C, and the functional relationship between the boost pump frequency and the return valve opening is obtained through the approximate model. A radial basis function neural network is selected, and the smoothness is adjusted to 0 to obtain the approximate model.

[0139] Specifically, in another embodiment, the functional relationship between the boost pump frequency and the opening of the return valve may be obtained through CFD simulation.

[0140] According to the functional relationship between the boost pump frequency and the return valve opening, the boost pump frequency and the return valve opening can be effectively adjusted to adjust the fuel flow returning to the buffer tank and control the temperature rise of the buffer tank, the boost pump and the fuel pipeline.

[0141] The method further comprises:

[0142] Acquire liquid level information in the first buffer tank 511 and the second buffer tank 512;

[0143] According to the liquid level information of the first buffer tank 511, the first liquid replenishing pump 321 is controlled to replenish liquid into the first buffer tank 511;

[0144] The second liquid replenishing pump 322 is controlled to replenish liquid into the second buffer tank 512 according to the liquid level information of the second buffer tank 512 .

[0145] Specifically, the method further includes:

[0146] Obtain the pressure difference data and fuel concentration data at the door of the explosion-proof room;

[0147] The explosion-proof fan is controlled to start, change speed or stop according to the pressure difference data at the explosion-proof room door and the fuel concentration data.

[0148] During the ignition process of the torch combustion system, the working pressure at the front end of the combustion nozzle is set to control the variable frequency output frequency of the booster pump to make the fuel burn more fully and reduce the amount of fuel drift.

[0149] It can be seen that in the above scheme, the present application obtains the output value of the pre-trained neural network model by inputting the obtained booster pump outlet pressure, fuel pipeline temperature, and burner nozzle front end pressure into the pre-trained neural network model: the three adjustable parameters Kp, Ki, and Kd of the PID controller, and timely adjustment of the booster pump frequency is achieved according to the output value, thereby improving the automatic control of the flare combustion and reducing energy waste. It should be understood that the size of the sequence number of each step in the above embodiment does not mean the order of execution, and the execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiment of the present invention.

[0150] In one embodiment, a flare combustion control device is provided, which corresponds one-to-one to the flare combustion control method in the above embodiment. Figure 7 As shown, the flare combustion control device includes an acquisition module 101, a neural network module 102, and an adjustment module 103. The detailed description of each functional module is as follows:

[0151] The acquisition module 101 is used to acquire the boost pump outlet pressure, the fuel pipeline temperature, and the burner nozzle front end pressure; specifically, the acquisition module 101 can be the PLC module group 220 in the first embodiment.

[0152] The neural network module 102 is used to input the boost pump outlet pressure, the fuel pipeline temperature, and the burner nozzle front end pressure as input values ​​into a pre-trained neural network model, and the output values ​​of the pre-trained neural network model are three adjustable parameters Kp, Ki, and Kd of the PID controller.

[0153] The adjustment module 103 is used to adjust the boost pump frequency according to the three adjustable parameters Kp, Ki, and Kd of the PID controller, and can be set in the MCGS configuration monitoring system 100 in the first embodiment.

[0154] The adjustment module 103 is further used to:

[0155] In the nth cycle, the return liquid in the burner is controlled to flow back to the first buffer tank 511 through the return liquid pipeline, and the second buffer tank 512 is replenished with liquid, and the fuel in the second buffer tank 512 is sent to the burner through the booster pump and the liquid supply pipeline;

[0156] In the n+1th cycle, the return liquid in the burner is controlled to flow back to the second buffer tank 512 through the return liquid pipeline, and the first buffer tank 511 is replenished with liquid, and the fuel in the first buffer tank 511 is sent to the burner through the booster pump and the liquid supply pipeline.

[0157] The acquisition module 101 is further used to:

[0158] Obtain the first temperature rise data of the fuel pipeline when the boost pump has different frequencies and the return valve has the same opening;

[0159] Obtain the second temperature rise data of the fuel pipeline when the boost pump has the same frequency and the return valve has different openings;

[0160] A flare combustion control device further includes a fitting module 104, which is used to:

[0161] According to the first temperature rise data of the fuel pipeline and the second temperature rise data of the fuel pipeline, a functional relationship between the frequency of the boost pump and the opening of the return valve is obtained.

[0162] The adjustment module 103 is further used to:

[0163] The frequency of the boost pump and the opening of the return valve are adjusted according to the functional relationship between the frequency of the boost pump and the opening of the return valve.

[0164] The acquisition module 101 is further used to:

[0165] Acquire liquid level information in the first buffer tank 511 and the second buffer tank 512;

[0166] The adjustment module 103 is further used to:

[0167] According to the liquid level information of the first buffer tank 511, the first liquid replenishing pump 321 is controlled to replenish liquid into the first buffer tank 511;

[0168] The second liquid replenishing pump 322 is controlled to replenish liquid into the second buffer tank 512 according to the liquid level information of the second buffer tank 512 .

[0169] The acquisition module 101 is further used to:

[0170] Obtain the pressure difference data and fuel concentration data at the door of the explosion-proof room;

[0171] The adjustment module 103 is further used to:

[0172] The explosion-proof fan is controlled to start, change speed or stop according to the pressure difference data at the explosion-proof room door and the fuel concentration data.

[0173] A flare combustion control device further includes a training module 105, which is used to:

[0174] Get Lemur algorithm parameters;

[0175] A matrix consisting of booster pump outlet pressure training data, fuel pipeline temperature training data, burner nozzle front end pressure training data, weight initial value and threshold initial value is used as the lemur population set;

[0176] Determine an initial position for starting iteration, and calculate an objective function for each lemur in the lemur population set in each iteration, wherein the objective function is the performance of the neural network model under current weights and thresholds;

[0177] According to the Lemur algorithm parameters, the free risk rate is calculated in each iteration;

[0178] According to the objective function of each lemur and the free risk rate in each iteration, updating the global optimal lemur and the nearest lemur of each lemur;

[0179] Obtain a random number, and update the lemur population set according to the random number and the free risk rate;

[0180] Based on the preset output conditions, the weights and thresholds represented by the global best lemur are output.

[0181] It is also used to: obtain a jump rate, and when the random number is greater than or equal to the jump rate, output the weight and threshold represented by the global best lemur.

[0182] The present invention provides a flare combustion control device. The present application obtains the booster pump outlet pressure, fuel pipeline temperature, and burner nozzle front end pressure into a pre-trained neural network model to obtain the output value of the pre-trained neural network model: three adjustable parameters Kp, Ki, and Kd of the PID controller. According to the output value, the booster pump frequency is adjusted in time, the automatic control of the flare combustion is improved, and the energy waste is reduced.

[0183] The specific definition of the flare combustion control device can be found in the definition of the flare combustion control method above, which will not be repeated here. Each module in the above-mentioned flare combustion control device can be implemented in whole or in part by software, hardware and a combination thereof. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0184] In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as follows: Figure 8 As shown. The computer device includes a processor, a memory, a network interface and a database connected via a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile and / or volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external device end through a network connection. When the computer program is executed by the processor, it realizes the functions or steps of a flare combustion control method service end side.

[0185] In one embodiment, a computer device is provided. The computer device may be a device end, and its internal structure diagram may be as follows: Fig. 9 As shown. The computer device includes a processor, a memory, a network interface, a display screen and an input device connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external server through a network connection. When the computer program is executed by the processor, it realizes the functions or steps of a flare combustion control method device side.

[0186] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the following steps when executing the computer program:

[0187] Obtain booster pump outlet pressure, fuel pipeline temperature, and burner nozzle front end pressure;

[0188] The boost pump outlet pressure, the fuel pipeline temperature, and the burner nozzle front end pressure are input as input values ​​into a pre-trained neural network model, and the output values ​​of the pre-trained neural network model are three adjustable parameters Kp, Ki, and Kd of the PID controller; wherein, when training the neural network model, the weight and threshold of the neural network model are updated using the Lemur algorithm;

[0189] The frequency of the boost pump is adjusted according to the three adjustable parameters Kp, Ki, and Kd of the PID controller.

[0190] In one embodiment, a computer readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented:

[0191] Obtain booster pump outlet pressure, fuel pipeline temperature, and burner nozzle front end pressure;

[0192] The boost pump outlet pressure, the fuel pipeline temperature, and the burner nozzle front end pressure are input as input values ​​into a pre-trained neural network model, and the output values ​​of the pre-trained neural network model are three adjustable parameters Kp, Ki, and Kd of the PID controller; wherein, when training the neural network model, the weight and threshold of the neural network model are updated using the Lemur algorithm;

[0193] The frequency of the boost pump is adjusted according to the three adjustable parameters Kp, Ki, and Kd of the PID controller.

[0194] It should be noted that the above functions or steps that can be implemented by the computer-readable storage medium or computer device can refer to the relevant descriptions on the server side and the device side in the aforementioned method embodiment. To avoid repetition, they will not be described one by one here.

[0195] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0196] Those skilled in the art can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example. In actual applications, the above-mentioned functions can be distributed and completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0197] The embodiments described above are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included in the protection scope of the present invention.

Claims

1. A flare combustion control method, characterized in that: include: Obtain booster pump outlet pressure, fuel pipeline temperature, and burner nozzle front end pressure; The boost pump outlet pressure, the fuel pipeline temperature, and the burner nozzle front end pressure are input as input values ​​into a pre-trained neural network model, and the output values ​​of the pre-trained neural network model are three adjustable parameters Kp, Ki, and Kd of a PID controller; The frequency of the boost pump is adjusted according to the three adjustable parameters Kp, Ki, and Kd of the PID controller.

2. A flare combustion control method according to claim 1, characterized in that: The method comprises: When training the neural network model, the weights and thresholds of the neural network model are updated using the Lemur algorithm, and the updating process includes: Get Lemur algorithm parameters; A matrix consisting of booster pump outlet pressure training data, fuel pipeline temperature training data, burner nozzle front end pressure training data, weight initial value and threshold initial value is used as the lemur population set; Determine an initial position for starting iteration, and calculate an objective function for each lemur in the lemur population set in each iteration, wherein the objective function is the performance of the neural network model under current weights and thresholds; According to the Lemur algorithm parameters, the free risk rate is calculated in each iteration; According to the objective function of each lemur and the free risk rate in each iteration, updating the global optimal lemur and the nearest lemur of each lemur; Obtain a random number, and update the lemur population set according to the random number and the free risk rate; Based on the preset output conditions, the weights and thresholds represented by the global best lemur are output.

3. A method for controlling flare combustion according to claim 2, characterized in that: The method of outputting the weight and threshold represented by the global best lemur based on the preset output condition includes: The jump rate is obtained, and when the random number is greater than or equal to the jump rate, the weight and threshold represented by the global best lemur are output.

4. A method for controlling flare combustion according to claim 1, characterized in that: The method further comprises: In the nth cycle, the return liquid in the burner is controlled to flow back to the first buffer tank through the return liquid pipeline, and the second buffer tank is replenished with liquid, and the fuel in the second buffer tank is sent to the burner through the booster pump and the liquid supply pipeline; In the (n+1)th cycle, the return liquid in the burner is controlled to flow back to the second buffer tank through the return liquid pipeline, and the first buffer tank is replenished with liquid, and the fuel in the first buffer tank is sent to the burner through the booster pump and the liquid supply pipeline.

5. A method for controlling flare combustion according to claim 2, characterized in that: The method further comprises: Obtain the first temperature rise data of the fuel pipeline when the boost pump has different frequencies and the return valve has the same opening; Obtain the second temperature rise data of the fuel pipeline when the boost pump has the same frequency and the return valve has different openings; Obtaining a functional relationship between a boost pump frequency and a return valve opening according to the first temperature rise data of the fuel pipeline and the second temperature rise data of the fuel pipeline; The frequency of the boost pump and the opening of the return valve are adjusted according to the functional relationship between the frequency of the boost pump and the opening of the return valve.

6. A flare combustion control method according to claim 1, characterized in that: The method further comprises: Acquire liquid level information in the first buffer tank and the second buffer tank; Controlling a first liquid replenishing pump to replenish liquid into the first buffer tank according to the liquid level information of the first buffer tank; According to the liquid level information of the second buffer tank, the second liquid replenishing pump is controlled to replenish liquid into the second buffer tank.

7. A flare combustion control method according to claim 1, characterized in that: The method further comprises: Obtain the pressure difference data and fuel concentration data at the door of the explosion-proof room; The explosion-proof fan is controlled to start, change speed or stop according to the pressure difference data at the explosion-proof room door and the fuel concentration data.

8. A flare combustion control device, characterized in that: include: An acquisition module is used to obtain the boost pump outlet pressure, fuel pipeline temperature, and burner nozzle front end pressure; A neural network module, used for inputting the boost pump outlet pressure, the fuel pipeline temperature, and the burner nozzle front end pressure as input values ​​into a pre-trained neural network model, wherein the output values ​​of the pre-trained neural network model are three adjustable parameters Kp, Ki, and Kd of a PID controller; The adjustment module is used to adjust the boost pump frequency according to the three adjustable parameters Kp, Ki, and Kd of the PID controller.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the flare combustion control method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the flare combustion control method according to any one of claims 1 to 7 are implemented.

Images