Closed cooling tower control method based on differential evolution algorithm

Abstract

The present invention provides a control method for a closed cooling tower based on a differential evolution algorithm, which relates to the technical field of closed cooling tower control. For the spray water pumps in a multi-closed cooling tower control system, the present invention introduces partial variable frequency regulation to replace on-off control, realizes variable frequency control of a single tower or multiple towers, establishes a cooling tower mechanism model and corrects parameters according to real-time data, comprehensively considers actual parameters, increases the target weight of the cooling tower outlet temperature when the ambient temperature is higher than the set threshold, and increases the target weights of the actuator action switching cost and operation energy consumption when the ambient temperature is lower than the set threshold, uses the differential evolution algorithm for operation and solution to obtain the optimized decision variables, and at the same time establishes an industrial closed-loop control communication link between the controller and the algorithm, collects on-site data through the controller and transmits it to the control room, imports the data into the algorithm in real time to obtain the decision variables, and issues the decision variables to the controller to effectively control the fans and spray pumps of each cooling tower.

Description

Technical Field

[0001] The present invention relates to the technical field of closed cooling tower control, and particularly relates to a closed cooling tower control method based on a differential evolution algorithm. Background Technique

[0002] A closed cooling tower is a common cooling device in industrial production. A heat exchanger is installed inside the tower. The process fluid to be cooled in industrial production enters the heat exchanger and dissipates heat through a spraying system and a ventilation system. At present, for the temperature control system of multiple closed cooling towers put into production earlier, due to hardware limitations, the spraying system mostly adopts on-off control, and the system control strategy adopts rule-based control, and there is great room for optimization in both hardware and software.

[0003] In the existing closed cooling tower control system, the spraying system mostly adopts on-off control, resulting in the inability to adjust the size of the spraying water volume and generating energy consumption. For the optimization method of the control strategy, the selected objective function is only the operating energy consumption of the actuator, without considering the actuator action switching cost and the ambient temperature. The judgment condition for optimization is based on the set error, and the selection of the error value has poor portability. Summary of the Invention

[0004] Aiming at the deficiencies of the prior art, the present invention provides a closed cooling tower control method based on a differential evolution algorithm to control the temperature of multiple closed cooling towers.

[0005] A closed cooling tower control method based on a differential evolution algorithm includes the following steps:

[0006] Step 1: Establish a mechanism model of a single closed cooling tower;

[0007] According to the principle of energy conservation, establish an energy balance equation among the circulating water in the pipe, the spraying water of the spraying pump, and the air.

[0008] The energy balance equation of the circulating water in the pipe is as follows:

[0009] M w C u dt = K0(t - t p )dA

[0010] In the formula, dA is the coil heat transfer area corresponding to the micro-element height segment H, M w is the mass flow rate of the cooling water, C w is the specific heat of the cooling water, K0 is the comprehensive heat transfer coefficient from the inside of the coil to the spraying water based on the outer surface of the pipe, t p is the temperature of the spraying water outside the pipe, and t is the temperature of the cooling water inside the pipe.

[0011] The energy balance equation of the spraying water of the spraying pump is as follows:

[0012] M p C p dt p =-K m (h p -h0)dA + K0(t - t p )dA

[0013] In the formula, M p is the mass flow rate of the spray water, C p is the specific heat of the spray water, h p is the enthalpy value of the saturated humid air corresponding to the temperature of the spray water, h0 is the enthalpy value of the inlet air, and K m is the mass transfer coefficient;

[0014] The energy balance equation of the air is as follows:

[0015] G a dh = K m (h p -h0)dA

[0016] In the formula, G a is the mass flow rate of the air, and h is the enthalpy value of the air;

[0017] Step 2: Solve the mechanism model of a single closed cooling tower to obtain the expression of the relationship between the outlet temperature of the circulating water, the spray water volume, the fan air volume, the ambient temperature, the air enthalpy value, and the heat transfer area of the cooling tower:

[0018] T out = f(t p , b1, b2, b3, b4, ψ1, ψ2, t1, h0, h1, h p , A0)

[0019] In the formula, t p is the spray water temperature, t1 is the inlet temperature of the cooling tower, h0 = f 01 (T g ), h1 = f 02 (T g ), h p = f 03 (T g ) are the inlet air enthalpy value, the outlet air enthalpy value, and the enthalpy value of the saturated humid air corresponding to the spray water temperature respectively, f 01 , f 02 , f 03 are expressed as functions of the ambient temperature T g , M p is the mass flow rate of the spray water, G a is the fan air volume, T g is the actual ambient wet bulb temperature, A0 is the heat transfer area of the cooling tower, and ψ1, ψ2 are the equations ψ 2The roots of +(b1 + b4)ψ+(b1b4 - b2b3)=0; b1 = f1(M p ); b2 = f2(M p ); b3 = f3(M p ); b4 = f4(G a ); f1, f2, f3 are expressed as functions of the spray water mass flow rate M p ; f4 is expressed as a function of the fan air volume G a .

[0020] Step 2.1: Let Substitute into the single closed-circuit cooling tower mechanism model:

[0021]

[0022] Step 2.2: Assume that the enthalpy value h p of saturated moist air has a linear relationship with the temperature of the spray water outside the pipe, that is: Let y = t p -t, z = h0 - h p ;

[0023] Step 2.3: Substitute the formula in Step 2.2 into 2.1 to get

[0024] where b1=(a1 + a3); b2=-a2; b3=-ma3; b4=ma2 - a4

[0025] Solve the system of equations:

[0026]

[0027] where ψ1, ψ2 are the roots of the equation ψ 2 +(b1 + b4)ψ+(b1b4 - b2b3)=0, m1, m2 are constant coefficients;

[0028] Step 2.4: Establish the boundary conditions of the system of equations in Step 2.3:

[0029] At the top of the cooling tower, at the cooling water inlet:

[0030] A = A0; t = t1; t p = t p0 = t p1 ; h = h1; h p = h p0 = h p1 ;

[0031]

[0032] where t1 is the cooling water inlet temperature, t p0, t p1 are the outlet temperature and the inlet temperature of the circulating water respectively, h1 is the air enthalpy value at the top of the cooling tower, h p0 , h p1 are the outlet enthalpy value and the inlet enthalpy value of the circulating water respectively.

[0033] At the bottom of the cooling tower, at the outlet of the cooling water, the boundary conditions are:

[0034] A = 0; t = t0; t p = t p0 = t p1 ; h = h0; h p = h p0 = h p1 ;

[0035] (t p0 - t0) = m1 + m2

[0036]

[0037] where t0 is the outlet temperature of the cooling water and h0 is the air enthalpy value at the bottom of the cooling tower.

[0038] Step 2.5: Calculate respectively from the boundary conditions at the inlet and outlet of the cooling water:

[0039]

[0040]

[0041] Step 2.6: Assume that the temperature of the circulating water remains unchanged, i.e., t p = t p0 = t p1 , there is:

[0042]

[0043] T out = (t 01 + t 02 ) / 2

[0044] where, t 01 , t 02 are the two obtained outlet temperature values of the cooling water, and T out is the outlet temperature of the cooling water.

[0045] Step 3: Determine the parameters of the mechanism model of the closed cooling tower according to the structural parameters of the closed cooling tower, and calibrate the parameters of the mechanism model of the closed cooling tower through the actual operation data;

[0046] The actual operation data includes:

[0047] The inlet temperature, outlet temperature, and ambient temperature. The cooling capacity of each cooling tower is differentiated by adjusting the heat transfer area, heat transfer coefficient, and mass transfer coefficient.

[0048] Step 3.1: Collect the structural parameters of the closed cooling tower, specifically including the cooling capacity, cooling load, inlet temperature, outlet temperature, spray water volume, air volume entering the tower, wet bulb temperature, dry bulb temperature, air relative humidity, enthalpy value of air entering the tower, enthalpy value of air leaving the tower, heat transfer coefficient, mass transfer coefficient, saturated water vapor partial pressure, heat transfer area, i.e., the pipe length, inner diameter, outer diameter, pipe spacing, number of pipes, and number of rows of the circulating water coil.

[0049] Step 3.2: Calculate the heat and mass transfer coefficients and the heat transfer area:

[0050] Among them, the heat transfer coefficient K0:

[0051]

[0052] In the formula, α i is the heat transfer coefficient of the fluid inside the coil, d i is the outer diameter of the pipe, λ is the thermal conductivity, α w is the convective heat transfer coefficient of the water film, δ is the coil thickness, λ g is the thermal conductivity of the water scale, D n , D c are the inner and outer diameters of the coil after scaling respectively;

[0053] The mass transfer coefficient K m :

[0054]

[0055] G max is the maximum value of the air mass flow rate, a′ is the convective heat transfer coefficient between the spray water and the air-water interface, and e is the proportionality coefficient;

[0056] The heat transfer area:

[0057] A = N′ p ·N·π·d0·L

[0058] N is the number of pipes in a row, N p is the number of pipe rows, d0 is the outer diameter of the pipe, and L is the pipe length;

[0059] Step 3.3: Select the actual operating data according to the meteorological conditions of the fixed area, including:

[0060] Atmospheric pressure: P = 101325 Pa

[0061] Saturated water vapor partial pressure: P s = 5535 Pa

[0062] Relative humidity:

[0063] Moisture content:

[0064] Enthalpy value of inlet air: h0 = 1.01T g +0.001d(2501 + 1.85T g )

[0065] Enthalpy value of outlet air:

[0066] Estimate the spray water temperature and its change, and estimate the parameter: T p

[0067] Step 3.4: Actually measure and input real-time variables, specifically including: ambient temperature T g , inlet temperature T i and outlet temperature T o ; Substitute the actually measured real-time data into the model to correct the mechanism model parameters;

[0068] Step 4: Introduce frequency conversion regulation at the spray pump of a single cooling tower;

[0069] Step 4.1: Analyze the influence of frequency on the rotational speed of the spray pump motor:

[0070] The rotational speed n of an asynchronous induction motor has the following relationship with three parameters: power supply frequency f, slip ratio s, and number of pole pairs p of the motor:

[0071]

[0072] Step 4.2: Analyze the relationship between the spray water volume and the rotational speed of the motor:

[0073] According to the principle of fluid mechanics, the flow rate of the water pump is linearly proportional to the rotational speed, and the power of the pump is proportional to the cube of the rotational speed. The power consumption of the motor using rotational speed regulation is:

[0074]

[0075] where n is the rotational speed, n0 is the rotational speed under rated conditions, N represents power, and N0 is the rated power under rated conditions;

[0076] Step 4.3: Select a matching frequency converter according to the actual closed cooling tower;

[0077] Step 5: Establish 4 optimized control models for closed cooling towers according to the mechanism model of a single closed cooling tower;

[0078] Step 5.1: Establish four objective functions for the optimal control model of the closed cooling tower, which are mainly based on the outlet temperature of the cooling tower, the switching frequency of the actuator, and the operating power of the actuator. The four closed cooling towers are set as Tower 1 to Tower 4;

[0079] J min = min λ1J a + λ2J b + λ3J c

[0080]

[0081] In the formula: J min is the objective function, J a is the outlet temperature of the cooling tower following the set temperature, J b is the switching frequency of the fan and the spray pump actions, J c is the operating power of the fan and the spray pump, λ1, λ2, λ3 are the weight coefficients of each state constraint, ξ1, ξ2, ξ3 are the weight coefficients of the switching frequency of the fan and the spray pump, is the outlet temperature of the cooling tower model, y set is the set value of the outlet temperature of the cooling tower, is the control state variable of the fan; is the control state variable of the spray pump; T min 、T max are the minimum and maximum values of the set outlet temperature of the cooling tower;

[0082] ε is a very small positive value, is the percentage of the spray water flow of the spray pump of Tower 4;

[0083] Step 5.2: Set the constraint conditions of the objective function as:

[0084]

[0085] Step 5.3: Adjust the objective weights according to the ambient temperature:

[0086] When the ambient temperature is higher than the set threshold, increase the objective weight λ1 of the outlet temperature of the cooling tower. When the ambient temperature is lower than the set threshold, increase the objective weights λ2 and λ3 of the actuator action switching cost and the operating energy consumption.

[0087] Step 6: According to the optimization problem, perform real number coding on the decision variables. The dimension of the decision variables is 8, which are the fan operating states and the spray pump operating states of the four cooling towers respectively. Use the differential evolution algorithm to solve the optimization problem;

[0088] Step 6.1: Decision variable coding: For the optimization problem The dimension D of the decision variable is 8.

[0089] X m = [x f1 , x f2 , x f3 , x f4 , x p1 , x p2 , x p3 , x p4

[0090] where x f1 , x f2 , x f3 , x f4 are the operating states of the fans of 4 towers respectively x p1 , x p2 , x p3 , x p4 are the operating states of the spray pumps of 4 towers respectively

[0091] Step 6.2: Parameter initialization: Set the maximum number of iterations g max , the iteration index g, and generate an initial population P containing NP individuals g = {x 1，g , x 2，g , … x NP，g}, where x i，g = {x 1 i，g , … x D i，g}, and each element is uniformly distributed in the interval [L min , L max = [0, 1], and D is the dimension of the decision variable;

[0092] Step 6.3: Perform differential iteration: Perform the mutation operation,

[0093]

[0094] i, r1, r2, r3 ∈ [1, NP], i ≠ r1 ≠ r2 ≠ r3, F ∈ [0, 1]

[0095] Take F = 0.5 as the mutation probability, and i, r1, r2, r3 are the numbers of different individuals in the population respectively.

[0096] Step 6.4: Crossover: Perform crossover for each individual and its generated offspring individuals,

[0097]

[0098] where CR is the crossover probability, which is a value between [0, 1]; ​

[0099] Step 6.5: Decoding: For the optimized decision variable x = [x f1 , x f2 , x f3 , x f4 , x p1 , x p2 , x p3 , x p4 , perform decoding. For x f1 , x f2 , x f3 , x f4 :

[0100]

[0101] For x p1 , x p2 , x p3 :

[0102]

[0103] For x p4 :

[0104]

[0105] Determine the control quantity according to the decoded decision variable;

[0106] Step 6.5: Calculate the value of the fitness function:

[0107] J min = min λ1J a + λ2J b + λ3J c

[0108]

[0109] Step 6.6: According to the value of the fitness function, select the better one from the target individuals and trial individuals as the next generation:

[0110]

[0111] Step 6.7: Store the decision variable and output it to the database table, and then send it to the PLC for control through the configuration software.

[0112] Step 7: Build the communication link between the controller and the algorithm, collect the real-time data of the industrial site through the controller, including the inlet temperature and outlet temperature of the circulating water of each cooling tower, and transmit the data to the monitoring room through the Ethernet. Taking the configuration software and the database as the interface, realize the import of the real-time data into the algorithm to obtain the control instructions for the fan and the spray pump, and then send them to the controller instructions for control;

[0113] Step 7.1: The configuration software first performs IO configuration on the controller through TCP / IP. After configuring the ODBC data source, it calls the built-in ADO component to read the data in the controller's data storage area in real time and communicate with the database, including the circulating water inlet temperature, circulating water outlet temperature, ambient temperature, spray water temperature, air enthalpy value, and the on / off states of each actuator of each cooling tower.

[0114] Step 7.2: The algorithm is written in Python language and calls pymssql to communicate with the database sqlserver. The database tables include the data table read from the controller and the control quantity data table obtained after algorithm operation, realizing the closed-loop control of the industrial field controller.

[0115] The beneficial effects produced by adopting the above technical solutions are as follows:

[0116] The present invention provides a closed cooling tower control method based on the differential evolution algorithm, having the following beneficial effects:

[0117] 1. For the spray water pumps in the multi-closed cooling tower control system, partial frequency conversion regulation is introduced to replace the on / off control, realizing the frequency conversion control of a single tower or multiple towers, which not only saves costs but also reduces energy consumption, and reduces the current impact during motor startup, prolonging the service life;

[0118] 2. A cooling tower mechanism model is established and parameter correction is performed according to real-time data, an optimal control model is established, and the differential evolution algorithm in the intelligent optimization algorithm is used for operation and solution to obtain the optimized decision variables, reducing the actuator on / off switching cost and operation energy consumption, and effectively controlling each cooling tower fan and spray pump;

[0119] 3. A communication link is established between the controller and the algorithm to realize the industrial closed-loop control communication link for importing real-time data into the algorithm and sending the decision command to the controller for control. BRIEF DESCRIPTION OF THE DRAWINGS

[0120] Figure 1 It is a schematic diagram of a single closed cooling tower control system in an embodiment of the present invention.

[0121] In the figure, 1 - spray water pump; 2 - fan; 3 - spray water; 4 - single tower circulating water inlet; 5 - single tower circulating water outlet; 6 - water tank; 7 - air inlet; 8 - air outlet;

[0122] Figure 2 It is a schematic diagram of the optimized control models of 4 closed cooling towers of the present invention.

[0123] In the figure, 9 - Circulating water inlet; 10 - Circulating water outlet; 11 - Frequency converter; 12 - Closed cooling tower 1#; 13 - Closed cooling tower 2#; 14 - Closed cooling tower 3#; 15 - Closed cooling tower 4#;

[0124] Figure 3 It is a schematic diagram of the control method for the closed cooling tower in the embodiment of the present invention;

[0125] Figure 4 It is a schematic diagram of the communication between the optimization algorithm of the closed cooling tower control method of the present invention and the controller;

[0126] Figure 5 It is a schematic diagram of the control result based on the differential evolution algorithm of the present invention. Detailed implementation manners

[0127] The following combines the accompanying drawings and embodiments to further describe in detail the specific implementation manners of the present invention. The following embodiments are used to illustrate the present invention, but are not used to limit the scope of the present invention.

[0128] With the increasingly intensifying climate change, the demand for cooling in industrial production is also increasing. As a modern cooling device, the cooling tower has more advantages than traditional air - conditioning systems and is a common cooling device in industrial production. Upgrading the hardware of existing cooling devices and optimizing the control methods can not only save costs but also reduce the energy consumption caused by the original control methods.

[0129] A heat exchanger is installed in the closed cooling tower. The process fluid to be cooled in industrial production enters the heat exchanger and dissipates heat through the spray system and ventilation system. For the control system of early - put - into - production closed cooling towers, the spray system of the cooling tower mostly adopts on - off control. If the spray water volume can be adjusted, the production energy consumption will be greatly reduced. With the update of frequency - conversion technology, frequency - conversion regulation can be introduced at the spray pump to achieve variable - frequency control of the spray water, with significant energy - saving effects, and it can also reduce the current impact during motor startup and extend the service life of the motor.

[0130] For the original cooling tower control system, a rule - based control method is mostly adopted, that is, specific control strategies are executed within a set temperature range. This method has a simple control logic, but it is difficult to achieve precise control. It does not consider environmental factors, the switching cost of actuator actions, and operating energy consumption. It is necessary to update the control strategy of the cooling tower. A cooling model is established based on parameters such as the cooling water volume and cooling area of the cooling tower, and the differential evolution algorithm is used to calculate the environmental wet - bulb and dry - bulb temperatures, inlet temperature, etc. collected by the controller, and the optimal solution is selected to control the cooling tower, which can achieve the production goal of high - efficiency energy saving.

[0131] A control method for a closed cooling tower based on the differential evolution algorithm, as Figure 3 shown, includes the following steps:

[0132] Step 1: Establish a mechanism model of a single closed cooling tower, as Figure 1 shown;

[0133] According to the principle of energy conservation, establish the energy balance equations among the circulating water in the pipe, the sprayed water by the spray pump, and the air;

[0134] The heat loss of the circulating water in the pipe is equal to the heat exchange between the circulating water and the sprayed water per unit area;

[0135] The increase in the heat of the sprayed water by the spray pump is equal to the heat exchange between the circulating water and the sprayed water per unit area, and the decrease in heat is equal to the energy loss of the sprayed water evaporated into the air;

[0136] The increase in the heat of the air is equal to the energy change of the sprayed water evaporated into the air.

[0137] The energy balance equation of the circulating water in the pipe is as follows:

[0138] M w C w dt = K0(t - t p )dA

[0139] where dA is the coil heat transfer area corresponding to the micro-element height segment H, M w is the mass flow rate of the cooling water, C w is the specific heat of the cooling water, K0 is the overall heat transfer coefficient from the inside of the coil to the sprayed water based on the outer surface of the pipe, t p is the temperature of the sprayed water outside the pipe, and is the temperature of the cooling water inside the pipe.

[0140] The energy balance equation of the sprayed water by the spray pump is as follows:

[0141] M p C p dt p = -K m (h p - h0)dA + K0(t - t p )dA

[0142] where M p is the mass flow rate of the sprayed water, C p is the specific heat of the sprayed water, h p is the enthalpy value of the saturated wet air corresponding to the temperature of the sprayed water, h0 is the inlet air enthalpy value, and K m is the mass transfer coefficient;

[0143] The energy balance equation of the air is as follows:

[0144] G a dh = K m (h p - h0)dA

[0145] where G a is the mass flow rate of air, and h is the enthalpy value of air;

[0146] Step 2: Solve the mechanism model of a single closed cooling tower to obtain the expression of the relationship between the outlet temperature of the circulating water and the spray water volume, the fan air volume, the ambient temperature, the air enthalpy value, and the heat transfer area of the cooling tower:

[0147] T out = f(t p , b1, b2, b3, b4, ψ1, ψ2, t1, h0, h1, h p , A0)

[0148] where t p is the spray water temperature, t1 is the inlet temperature of the circulating water, h0 = f 01 (T g ), h1 = f 02 (T g ), h p = f 03 (T g ) are the inlet air enthalpy value, the outlet air enthalpy value, and the enthalpy value of saturated humid air corresponding to the spray water temperature respectively, f 01 , f 02 , f 03 are expressed as functions of the ambient temperature T g , M p is the mass flow rate of the spray water, G a is the fan air volume, T g is the actual ambient wet bulb temperature, A0 is the heat transfer area of the cooling tower, ψ1, ψ2 are the roots of the equation ψ 2 +(b1 + b4)ψ+(b1b4 - b2b3)= 0; b1 = f1(M p ); b2 = f2(M p ); b3 = f3(M p ); b4 = f4(G a ); f1, f2, f3 are expressed as functions of the mass flow rate of the spray water M p , and f4 is expressed as a function of the fan air volume G a .

[0149] Step 2.1: Let Substitute it into the mechanism model of a single closed cooling tower:

[0150]

[0151] Step 2.2: Assume that the enthalpy value h p of saturated humid air has a linear relationship with the temperature of the spray water outside the pipe, that is: Let y = tp -t, z = h0 - h p ;

[0152] Step 2.3: Substitute the formula in Step 2.2 into 2.1, we get

[0153] where b1 = (a1 + a3); b2 = -a2; b3 = -ma3; b4 = ma2 - a4

[0154] Solve the system of equations:

[0155]

[0156] where ψ1, ψ2 are the roots of the equation ψ 2 +(b1 + b4)ψ + (b1b4 - b2b3) = 0, m1, m2 are constant coefficients;

[0157] Step 2.4: Establish the boundary conditions for the system of equations in Step 2.3:

[0158] At the top of the cooling tower, at the cooling water inlet:

[0159] A = A0; t = t1; t p = t p0 = t p1 ; h = h1; h p = h p0 = h p1 ;

[0160]

[0161] where t1 is the cooling water inlet temperature, t p0 , t p1 are the circulating water outlet temperature and inlet temperature respectively, h1 is the air enthalpy value at the top of the cooling tower, h p0 , h p1 are the outlet enthalpy value and inlet enthalpy value of the circulating water respectively.

[0162] At the bottom of the cooling tower, at the cooling water outlet, the boundary conditions are:

[0163] A = 0; t = t0; t p = t p0 = t p1 ; h = b0; h p = h p0 = h p1 ;

[0164] (t p0 - t0) = m1 + m2

[0165]

[0166] Wherein, t0 is the outlet temperature of the cooling water, and h0 is the enthalpy value of the air at the lower end of the cooling tower.

[0167] Step 2.5: Calculate respectively from the boundary conditions at the inlet and outlet of the cooling water:

[0168]

[0169] Step 2.6: Assume that the temperature of the circulating water remains unchanged, i.e., t p = t p0 = t p1 , then we have:

[0170]

[0171] T out = (t 01 + t 02 ) / 2

[0172] Wherein, t 01 , t 02 are two obtained outlet temperature values of the cooling water, and T out is the outlet temperature of the cooling water.

[0173] Step 3: Determine the parameters of the mechanism model of the closed cooling tower according to the structural parameters of the closed cooling tower, and correct the parameters of the mechanism model of the closed cooling tower through the actual operation data;

[0174] The actual operation data includes:

[0175] The inlet temperature of the circulating water, the outlet temperature of the circulating water, the ambient temperature, and the difference in the cooling capacity of each cooling tower is distinguished by adjusting the heat transfer area, the heat transfer coefficient, and the mass transfer coefficient.

[0176] Step 3.1: Collect the structural parameters of the closed cooling tower, specifically including the cooling capacity, cooling load, inlet temperature of the circulating water, outlet temperature of the circulating water, spray water volume, air volume entering the tower, wet bulb temperature, dry bulb temperature, relative humidity of the air, enthalpy value of the air entering the tower, enthalpy value of the air leaving the tower, heat transfer coefficient, mass transfer coefficient, saturated water vapor partial pressure, heat transfer area, i.e., the pipe length, inner diameter, outer diameter, pipe spacing, number of pipes, and number of rows of the circulating water coil.

[0177] In this embodiment, the input structural parameters of the cooling tower are as follows:

[0178] The coil is a smooth stainless steel pipe with a thermal conductivity coefficient λ = 54 w / (m·k), an outer diameter d0 = 22 mm, an outer diameter d i = 20 mm, a pipe length L = 4 mm, a pipe spacing P t = 40 mm, and is arranged in a staggered equilateral triangle; the number of pipes in the row N = 58, and the number of pipe rows N p= 32. Cooling water flow rate G w = 100 (t / h), fixed parameter of the cooling tower, inlet temperature of circulating water T in = 43 °C, fixed parameter of the cooling tower, outlet temperature of circulating water T out = 33 °C, specific heat capacity of water c w = 4.19×10 3 , total heat transfer amount: Q = G w c w (T in - T out ), maximum air volume of the fan M p = 93 (t / h), maximum spray water volume G a = 174 (t / h)

[0179] Step 3.2: Calculate the heat and mass transfer coefficient and heat transfer area:

[0180] Among them, the heat transfer coefficient K0:

[0181]

[0182] In the formula, α i is the heat transfer coefficient of the fluid inside the coil, d i is the outer diameter of the pipe, λ is the thermal conductivity, α w is the convective heat transfer coefficient of the water film, δ is the thickness of the coil, λ g is the thermal conductivity of the water scale, D n , D c are the inner and outer diameters of the coil after scaling respectively; referring to the data, the initial reference value is K0 = 873.8 w(m 2 ·k), and it is corrected according to the actual data.

[0183] Mass transfer coefficient K m : According to the calculation formula proposed by Parker et al.

[0184]

[0185] G max is the maximum value of the air mass flow rate, a′ is the convective heat transfer coefficient between the spray water and the air-water interface, and e is the proportionality coefficient. Referring to the data, the initial reference value is K m = 0.28 kg(m 2 ·s), and it is corrected according to the actual data.

[0186] Heat transfer area:

[0187] A = N p ·N·π·d0·L

[0188] N is the number of rows of pipes, N pis the number of tube rows, d0 is the outer diameter of the tube, L is the tube length, and the initial reference value is A = 580m 2 , and it is corrected according to the actual data.

[0189] Step 3.3: Select the actual operating data according to the meteorological conditions in a fixed area, including:

[0190] Atmospheric pressure: P = 101325Pa

[0191] Saturated partial pressure of water vapor: P s = 5535Pa

[0192] Relative humidity:

[0193] Moisture content:

[0194] Enthalpy value of inlet air: h0 = 1.01T g + 0.001d(2501 + 1.85T g )

[0195] Enthalpy value of outlet air:

[0196] Estimate the spray water temperature and its change, and estimate the parameter: T p

[0197] Step 3.4: Actually measure the input real-time variables, specifically including: ambient temperature T g , inlet temperature of circulating water T i and outlet temperature of circulating water T o ; Substituting the actually measured real-time data into the model can correct the parameters of the mechanism model.

[0198] Step 4: Taking 4 parallel closed cooling towers as the background, in order to improve the operating efficiency of the cooling tower, considering energy conservation and cost comprehensively, variable frequency regulation is introduced at the spray pump of a single cooling tower;

[0199] The closed cooling tower adjusts the air flow by changing the operating frequency of the fan and adjusts the spray water volume by changing the operating frequency of the spray pump. The fan in a single cooling tower has 3 operating states: high speed, low speed, and shutdown. The high and low speed conversion is realized through the star-delta conversion of the circuit. The spray pump has 2 operating states, start and stop. After introducing variable frequency regulation, the spray pump realizes variable frequency regulation;

[0200] Step 4.1: Analyze the influence of frequency on the rotational speed of the spray pump motor:

[0201] The rotational speed n of an asynchronous induction motor has the following relationship with three parameters: power supply frequency f, slip ratio s, and number of pole pairs p of the motor:

[0202]

[0203] Changing the power supply frequency f can change the motor speed. Since the frequency converter has excellent speed regulation performance, using a variable frequency speed regulation device can make the spray water pump operate in the best state by changing the power supply frequency.

[0204] Step 4.2: Analyze the relationship between the spray water volume and the motor speed:

[0205] According to the principle of fluid mechanics, the flow rate of the water pump is linearly proportional to the speed, and the power of the pump is proportional to the cube of the speed. The power consumption of the motor using speed regulation is:

[0206]

[0207] where n is the speed, n0 is the rated working condition speed, N represents power, and N0 is the rated working condition power; when the flow rate is reduced to 0.8 and the speed is reduced to 0.8, the power of the motor is 51.2% of the rated power, saving 48.8% of the electric energy. The energy-saving effect is significant, and it can reduce the current impact during motor startup and extend the service life of the motor.

[0208] Step 4.3: Select a matching frequency converter according to the actual closed cooling tower:

[0209] In this embodiment, the spray water volume of the closed cooling tower is 100m 3 / h, the power of the water pump motor is 2.2kW, and a Rockwell frequency converter PowerFlex525 with a configured power of 2.2kW is used to realize the adjustment of the spray water volume.

[0210] Step 5: According to the mechanism model of a single closed cooling tower, establish 4 optimized control models for closed cooling towers, as Figure 2 shown;

[0211] For multiple closed cooling towers, the optimized control objective is to make the average temperature of the circulating water outlet of multiple cooling towers reach the set temperature to ensure the safe operation of production. At the same time, the energy consumption cost of the cooling tower within the control time domain is minimized; since the cooling capacity of the cooling tower can reach the set circulating water outlet temperature most of the time. To make the outlet temperature of the cooling tower tend to the set value in an optimal way at future sampling points, the objective function adopts the form of the absolute value between the predicted temperature and the reference value. When the set outlet temperature is reached, the switching frequency of the fan and the spray pump should be minimized, and the operating power of the fan and the spray pump should be reduced. The variable frequency adjustment of the spray pump of a single tower is introduced to establish an optimized objective function for the circulating water outlet temperature.

[0212] Step 5.1: Establish 4 optimized control model objective functions for closed cooling towers mainly based on the circulating water outlet temperature, actuator switching frequency, and actuator operating power, where the 4 closed cooling towers are set as Tower 1 to Tower 4;

[0213] J min = min λ1J a + λ2J b + λ3J c

[0214]

[0215] where: J min is the objective function, J a is the circulating water outlet temperature following the set temperature, J b is the action switching frequency of the fan and the spray pump, J c is the operating power of the fan and the spray pump, λ1, λ2, λ3 are the weight coefficients of each state constraint, changing this parameter can adjust the performance of the system; ξ1, ξ2, ξ3 are the weight coefficients of the switching frequency of the fan and the spray pump, changing this parameter can affect the actuator action result; is the circulating water outlet temperature of the cooling tower model, y set is the set value of the circulating water outlet temperature, is the fan control state variable; is the spray pump control state variable; T min T max are the minimum and maximum values of the set circulating water outlet temperature;

[0216] is a very small positive value, is the percentage of the spray water flow of the spray pump of Tower No. 4, to prevent the motor from overheating and burning due to insufficient heat dissipation during low-speed operation of the motor, take 0.5 - 1.

[0217] Step 5.2: Set the constraint conditions of the objective function as:

[0218]

[0219] Step 5.3: Adjust the objective weights according to the ambient temperature:

[0220] When the ambient temperature is higher than the set threshold, increase the objective weight λ1 of the circulating water outlet temperature of the cooling tower. When the ambient temperature is lower than the set threshold, increase the objective weights λ2, λ3 of the actuator action switching cost and the operating energy consumption.

[0221] Step 6: According to the optimization problem, perform real - number coding on the decision variables. The dimension of the decision variables is 8, which are the operating states of the fans and the spray pumps of 4 cooling towers respectively. Use the differential evolution algorithm to solve the optimization problem;

[0222] First, initialize the population, including the maximum number of iterations, the number of individuals in the population, and the upper and lower bounds of the decision variables. Second, perform differential evolution iteration. Use the classical mutation strategy DE / rand / 1 to generate a mutant individual. For each individual and its generated offspring individual, perform crossover. For the optimized decision variables obtained, perform decoding. According to the value of the fitness function, select the better one from the target individual and the trial individual as the next generation. Finally, store the decision variables and output them to the database table, and then send them to the PLC for control through the configuration software.

[0223] Step 6.1: Encoding of decision variables: For the optimization problem The dimension D of the decision variable is 8.

[0224] X m =[x f1 ,x f2 ,x f3 ,x f4 ,x p1 ,x p2 ,x p3 ,x p4

[0225] where x f1 ,x f2 ,x f3 ,x f4 are the fan operating states of 4 towers respectively x p1 ,x p2 ,x p3 ,x p4 are the spray pump operating states of 4 towers respectively Since the first 7 decision variables are integers and the 8th decision variable is a real number, adopt the real number encoding method to perform real number encoding on all decision variables.

[0226] Step 6.2: Parameter initialization: Set the maximum number of iterations g max =50, the iteration index g = 0, and generate an initial population P g ={x 1，g ,x 2，g ,…x NP，g}, where x i，g ={x 1 i，g ,…x D i，g}, and each element is uniformly distributed in the interval [L min ,U max =[0,1], D is the dimension of the decision variable;

[0227] ​Step 6.3: Perform differential iteration: Mutation: For the mutation strategy of the differential evolution algorithm, the base vector of the DE / rand / 1 method is randomly selected, which has a moderate convergence speed and population diversity. Therefore, this method is selected for the mutation operation.

[0228]

[0229] i, r1, r2, r3 ∈ [1, NP], i ≠ r1 ≠ r2 ≠ r3, F ∈ [0, 1]

[0230] Take F = 0.5 as the mutation probability, and i, r1, r2, r3 are the numbers of different individuals in the population respectively.

[0231] Step 6.4: Crossover: Perform crossover for each individual and its generated offspring individual.

[0232]

[0233] Among them, CR is the crossover probability, which is a value between [0, 1]. In this embodiment, CR = 0.7 is taken.

[0234] Step 6.5: Decoding: For the optimized decision variable x = [x f1 , x f2 , x f3 , x f4 , x p1 , x p2 , x p3 , x p4 , perform decoding. For x f1 , x f2 , x f3 , x f4 :

[0235]

[0236] For x p1 , x p2 , x p3 :

[0237]

[0238] For x p4 :

[0239]

[0240] If x = [0.54, 0.61, 0.56, 0.71, 1, 0.91, 0.99, 0.63], then the decision variable obtained is:

[0241] x = [1, 1, 1, 2, 1, 1, 1, 0.63]

[0242] Determine the control quantity according to the decoded decision variable; specifically: the fan of Tower 1 runs at low speed, the fan of Tower 2 runs at low speed, the fan of Tower 3 runs at low speed, the fan of Tower 4 runs at high speed, the spray pump of Tower 1 is turned on, the spray pump of Tower 2 is turned on, the spray pump of Tower 3 is turned on, and the flow rate of the spray pump of Tower 4 is 0.63.

[0243] Step 6.5: Calculate the value of the fitness function:

[0244] J min = min λ1J a + λ2J b + λ3J c

[0245]

[0246] Step 6.6: According to the value of the fitness function, select the better one from the target individuals and trial individuals as the next generation:

[0247]

[0248] Step 6.7: Store the decision variable and output it to the database table, and then send it to the PLC for control through the configuration software.

[0249] Step 7: Build a communication link between the controller and the algorithm, collect the real-time data of the industrial site through the controller, including the inlet temperature and outlet temperature of the circulating water of each cooling tower, and transmit the data to the monitoring room through Ethernet. Taking the configuration software and the database as the interface, realize the import of real-time data into the algorithm to obtain the control instructions for the fan and the spray pump, and then send them to the controller instructions for control, as Figure 4 shown;

[0250] Step 7.1: The configuration software first performs IO configuration on the controller through TCP / IP. After configuring the ODBC data source, call the built-in ADO component to communicate with the database by reading the data in the data storage area of the controller in real time, including the inlet temperature and outlet temperature of the circulating water of each cooling tower, the ambient temperature, the spray water temperature, the air enthalpy value, and the on / off states of each actuator.

[0251] Step 7.2: The algorithm is written in Python language, and pymssql is called to communicate with the database sqlserver. The database tables include the data table read from the controller and the control quantity data table obtained after algorithm operation; through the above industrial communication method, realize the closed-loop control of the intelligent optimization algorithm including the differential evolution algorithm for the industrial field controller, and the schematic diagram of the control result is as Figure 5 shown.

[0252] The above description is only a preferred embodiment of the present disclosure and an explanation of the applied technical principles. Those skilled in the art should understand that the scope of the invention involved in the embodiments of the present disclosure is not limited to the technical solutions formed by the specific combination of the above technical features, and should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the above inventive concept. For example, the technical solutions formed by mutually replacing the above features with the technical features (but not limited to) disclosed in the embodiments of the present disclosure that have similar functions.

Claims

1. A closed cooling tower control method based on differential evolution algorithm, characterized in that, It includes the following steps: Step 1: Establish a mechanism model of a single closed cooling tower; Step 2: Solve the mechanism model of the single closed cooling tower to obtain an expression for the relationship between the outlet temperature of the circulating water and the spray water volume, the air volume of the fan, the ambient temperature, the air enthalpy value, and the heat transfer area of the cooling tower; Step 3: Determine the parameters of the mechanism model of the closed cooling tower according to the structural parameters of the closed cooling tower, and calibrate the parameters of the mechanism model of the closed cooling tower through actual operation data; Step 4: Introduce frequency conversion regulation at the spray water pump of a single cooling tower; Step 5: Establish 4 optimization control models of closed cooling towers according to the mechanism model of a single closed cooling tower; Step 6: According to the optimization problem, perform real number coding on the decision variables. The dimension of the decision variables is 8, which are the operating states of the fans and the operating states of the spray pumps of 4 cooling towers respectively. Use the differential evolution algorithm to solve the optimization problem; Step 7: Build a communication link between the controller and the algorithm. Collect real-time data from the industrial site through the controller, including the inlet temperature and outlet temperature of the circulating water of each cooling tower, and transmit the data to the monitoring room through Ethernet. Use the configuration software and database as the interface to import the real-time data into the algorithm to obtain the control instructions for the fans and spray pumps, and then send them to the controller instructions for control.

2. The closed cooling tower control method based on differential evolution algorithm according to claim 1, characterized in that The specific content of Step 1 is: According to the principle of energy conservation, establish the energy balance equations among the circulating water in the pipe, the spray water of the spray pump, and the air; The energy balance equation of the circulating water in the pipe is as follows: M w C w dt = K0(t - t p )dA where dA is the coil heat transfer area corresponding to the differential height section H, M w is the mass flow rate of the cooling water, C w is the specific heat of the cooling water, K0 is the overall heat transfer coefficient from the inside of the coil to the spray water based on the outer surface of the pipe, t p is the spray water temperature, and t is the temperature of the cooling water inside the pipe; The energy balance equation of the spray water of the spray pump is as follows: M p C p dt p = -K m (h p -h0)dA + K0(t - t p )dA where M p is the mass flow rate of the spray water, C p is the specific heat of the spray water, h p is the enthalpy value of saturated humid air corresponding to the spray water temperature, h0 is the inlet air enthalpy value, K m is the mass transfer coefficient; The energy balance equation of the air is as follows: G a dh = K m (h p - h0)dA where G a is the mass flow rate of air and h is the enthalpy value of air.

3. A closed cooling tower control method based on differential evolution algorithm according to claim 1, characterized in that, The expression in Step 2 is: T out = f(t p , b1, b2, b3, b4, ψ1, ψ2, t1, h0, h1, h p , A0) where t p is the spray water temperature, t1 is the inlet temperature of the circulating water, h0 = f 01 (T g ), h1 = f 02 (T g ), h p = f 03 (T g ) are the inlet air enthalpy value, the outlet air enthalpy value, and the saturated moist air enthalpy value corresponding to the spray water temperature respectively, f 01 , f 02 , f 03 are expressed as functions of the ambient temperature T g , M g is the mass flow rate of the spray water, G a is the air volume of the fan, T g is the actual ambient wet bulb temperature, A0 is the heat transfer area of the cooling tower, ψ1, ψ2 are the roots of the equation ψ 2 +(b1 + b4)ψ+(b1b4 - b2b3)=0; b1 = f1(M p ); b2 = f2(M p ); b3 = f3(M p ); b4 = f4(G a ); f1, f2, f3 are expressed as functions of the mass flow rate M p of the spray water, and f4 is expressed as a function of the air volume G a of the fan.

4. A closed cooling tower control method based on differential evolution algorithm according to claim 1, characterized in that The specific content of Step 2 includes: Step 2.1: Set Substitute into the mechanism model of a single closed-circuit cooling tower: Step 2.2: Assume that the enthalpy value h of saturated moist air p has a linear relationship with the temperature of the sprayed water outside the pipe, i.e.: Let y = t p - t, z = h0 - h p ; Step 2.3: Substitute the formula in Step 2.2 into 2.1, and we get where b1 = (a1 + a3); b2 = -a2; b3 = -ma3; b4 = ma2 - a4 Solve the system of equations: where ψ1 and ψ2 are the roots of the equation ψ 2 +(b1 + b4)ψ+(b1b4 - b2b3)= 0, and m1 and m2 are constant coefficients; Step 2.4: Establish the boundary conditions of the system of equations in Step 2.3: At the top of the cooling tower, at the inlet of the cooling water: A = A0; t = t1; t p = t p0 = t p1 ; h = h1; h p = h p0 = h p1 ; Among them, t1 is the inlet temperature of the circulating water, t p0 , t p1 are the outlet temperature and the inlet temperature of the circulating water respectively, h1 is the air enthalpy value at the top of the cooling tower, h p0 , h p1 are the outlet enthalpy value and the inlet enthalpy value of the circulating water respectively; At the bottom of the cooling tower, at the outlet of the cooling water, the boundary conditions are: A = 0; t = t0; t p = t p0 = t p1 ; h = h0; h p = h p0 = h p1 ; (t p0 - t0) = m1 + m2 where t0 is the outlet temperature of the cooling water and h0 is the air enthalpy value at the lower end of the cooling tower; Step 2.5: Calculate respectively from the boundary conditions at the inlet and outlet of the cooling water: Step 2.6: Assume that the temperature of the circulating water remains unchanged, i.e., t p = t p0 = t p1 , then we have: T out = (t 01 + t 02 ) / 2 Among them, t 01 , t 02 are two obtained cooling water outlet temperature values, and T out is the cooling water outlet temperature.

5. A closed cooling tower control method based on differential evolution algorithm according to claim 1, characterized in that The actual operation data in Step 3 includes: The inlet temperature of the circulating water, the outlet temperature of the circulating water, the ambient temperature. The difference in the cooling capacity of each cooling tower is distinguished by adjusting the heat transfer area, heat transfer coefficient, and mass transfer coefficient.

6. A closed cooling tower control method based on differential evolution algorithm according to claim 1, characterized in that The specific content of Step 3 includes: Step 3.1: Collect the structural parameters of the closed cooling tower, specifically including the cooling capacity, cooling load, inlet temperature of the circulating water, outlet temperature of the circulating water, spray water volume, inlet air volume, wet bulb temperature, dry bulb temperature, air relative humidity, inlet air enthalpy value, outlet air enthalpy value, heat transfer coefficient, mass transfer coefficient, saturated water vapor partial pressure, heat transfer area, that is, the pipe length, inner diameter, outer diameter, pipe spacing, number of pipes, and number of rows of the circulating water coil; Step 3.2: Calculate the heat transfer and mass transfer coefficients and the heat transfer area: Among them, the heat transfer coefficient K0: where α i is the heat transfer coefficient of the fluid inside the coil, d i is the outer diameter of the pipe, λ is the thermal conductivity, α w is the convective heat transfer coefficient of the water film, δ is the thickness of the coil, λ g is the thermal conductivity of the scale, D n , D c are the inner and outer diameters of the coil after scaling, respectively; The mass transfer coefficient Kx: G max is the maximum value of the air quality flow velocity, a′ is the convective heat transfer coefficient between the sprayed water and the air-water interface, and e is the proportionality coefficient; The heat transfer area: A = N p ·N·π·d0·L N is the number of tubes arranged in a row, N p is the number of tube rows, d0 is the outer diameter of the tube, and L is the tube length; Step 3.3: Select actual operation data according to the meteorological conditions of a fixed area, including: Atmospheric pressure: P = 101325Pa Vapor saturation partial pressure: P s = 5535 Pa Relative humidity: Moisture content: Inlet air enthalpy value: h0 = 1.01T g + 0.001d(2501 + 1.85T g ) Outlet air enthalpy value: Estimate the spray water temperature and its changes, and estimate the parameter: p Step 3.4: Actually measure the input real-time variables, specifically including: ambient temperature T g , the inlet temperature T of the circulating water i and the outlet temperature T of the circulating water o ; Substitute the actually measured real-time data into the model to correct the mechanism model parameters.

7. A closed cooling tower control method based on differential evolution algorithm according to claim 1, characterized in that The specific steps of step 4 include: Step 4.1: Analyze the influence of frequency on the rotational speed of the spray pump motor: The rotational speed n of an asynchronous induction motor has the following relationship with three parameters: the power supply frequency f, the slip ratio s, and the number of pole pairs p of the motor: Step 4.2: Analyze the relationship between the spray water volume and the motor rotational speed: According to the principle of fluid mechanics, the flow rate of the water pump is linearly proportional to the rotational speed, and the power of the pump is proportional to the cube of the rotational speed. The power consumption of the motor using rotational speed regulation is: where n is the rotational speed, n0 is the rated operating condition rotational speed, N represents power, and N0 is the rated operating condition power; Step 4.3: Select a matching frequency converter according to the actual closed-circuit cooling tower.

8. A closed cooling tower control method based on a differential evolution algorithm according to claim 1, characterized in that The specific steps of step 5 include: Step 5.1: Establish four optimization control model objective functions mainly based on the circulating water outlet temperature, actuator switching frequency, and actuator operating power for the closed-circuit cooling tower. Among them, the four closed-circuit cooling towers are set as Tower 1 to Tower 4; J min = min λ1J a + λ2J b + λ3J c Where: J min is the objective function, J a is the circulating water outlet temperature following the set temperature, J b is the action switching frequency of the fan and the spray pump, J c is the operating power of the fan and the spray pump, λ1, λ2, and λ3 are the weight coefficients of each state constraint, and ξ1, ξ2, and ξ3 are the weight coefficients of the switching frequency of the fan and the spray pump, is the circulating water outlet temperature of the cooling tower model, y set is the set value of the circulating water outlet temperature, is the fan control state variable; is the spray pump control state variable; T min 、T max are the minimum and maximum values of the set circulating water outlet temperature; ε is a very small positive value, is the spray water flow percentage of the spray pump of Tower 4; Step 5.2: Set the constraint conditions of the objective function as: Step 5.3: Adjust the objective weight according to the ambient temperature: When the ambient temperature is higher than the set threshold, increase the objective weight λ1 of the circulating water outlet temperature. When the ambient temperature is lower than the set threshold, increase the objective weights λ2 and λ3 of the actuator action switching cost and operating energy consumption.

9. A closed cooling tower control method based on a differential evolution algorithm according to claim 1, characterized in that The specific steps of step 6 include: Step 6.1: Decision variable encoding: For the optimization problem The dimension D of the decision variable is 8; X m = [x f1 , x f2 , x f3 , x f4 , x p1 , x p2 , x p3 , x p4 ​ where x f1 , x f2 , x f3 , x f4 are the operating states of the fans of 4 towers respectively x p1 , x p2 , x p3 , x p4 are the operating states of the spray pumps of 4 towers respectively Step 6.2: Parameter initialization: Set the maximum number of iterations \(g\) max , the iteration index \(g\), and generate an initial population \(P\) containing \(NP\) individuals g =\(\{x\) 1，g， x 2，g , \(\cdots x\) NP，g \}\), where \(x\) i，g =\(\{x\) 1 i，g , \(\cdots x\) D i，g \}\), and each element is uniformly distributed in the interval \([L\) min , \(U\) max =[0, 1], and \(D\) is the dimension of the decision variable; Step 6.3: Perform differential iteration: Perform mutation operation, i, r1, r2, r3 ∈ [1, NP], i ≠ r1 ≠ r2 ≠ r3, F ∈ [0, 1] Take F = 0.5 as the mutation probability. i, r1, r2, and r3 are the numbers of different individuals in the population respectively; Step 6.4: Crossover: Perform crossover for each individual and its generated offspring individual, where CR is the crossover probability and is a value between [0, 1]; Step 6.5: Decoding: For the optimized decision variable x = [x f1 , x f2 , x f3 , x f4 , x p1 , x p2 , x p3 , x p4 , perform decoding. For x f1 , x f2 , x f3 , x f4 : For x p1 ,x p2 ,x p3 : For x p4 : Determine the control quantity according to the decoded decision variable; Step 6.5: Calculate the value of the fitness function: J min = min λ1J a + λ2J b + λ3J c Step 6.6: According to the value of the fitness function, select the better one from the target individual and the trial individual as the next generation: Step 6.7: Store the decision variable and output it to the database table, and then send it to the PLC for control through the configuration software.

10. A closed cooling tower control method based on a differential evolution algorithm according to claim 1, characterized in that, The specific steps of step 7 include: Step 7.1: The configuration software first performs IO configuration on the controller through TCP / IP. After configuring the ODBC data source, call the built-in ADO component to read the data in the data storage area of the controller in real time and communicate with the database, including the circulating water inlet temperature, circulating water outlet temperature, ambient temperature, spray water temperature, air enthalpy value, and switch status of each actuator of each cooling tower; Step 7.2: The algorithm is written in Python language, and pymssql is called to communicate with the database sqlserver. The database table includes the data table read from the controller and the control quantity data table obtained after algorithm operation, realizing the closed-loop control of the industrial field controller.

Images