Real-time optimization system for the working point of the new energy battery coating and drying control system

Abstract

This invention relates to a real-time optimization system for the operating point of a drying control system after coating of new energy batteries. It constructs a multi-objective optimization function based on energy consumption and NMP concentration indicators. By solving this optimization problem, the control target for NMP concentration, the frequency of the internal circulation fan, and the minimum frequency of the external circulation fan are determined. The objective function of the constructed optimization problem is achieved by solving a constrained linear optimization method. Each time the RTO optimization control module is invoked, the optimal NMP concentration control target, internal circulation fan target, and lower limit protection value of the external circulation fan frequency for the next calling cycle are obtained by solving the optimization problem. This achieves the goal of minimizing the total circulating air volume of the system while keeping the NMP concentration within the limit, thereby achieving energy-saving and economical operation.

Description

Technical Field

[0001] This invention belongs to the field of drying control after battery coating, and particularly relates to a real-time optimization system for the operating point of a drying control system after coating of new energy batteries. Background Technology

[0002] The post-coating drying process is a crucial step in new energy battery production. It involves drying the wet film after coating the battery. In the drying chamber, circulating high-temperature hot air is used to dry the coating. Because NMP solvent is used in the battery cathode coating, this solvent vaporizes during the drying process, producing a weakly toxic gas. When the gas concentration reaches a certain level, there is a risk of explosion. Therefore, the battery drying process must strictly control the NMP gas concentration within safe limits. The battery coating drying system employs a cascaded design of multiple ovens. The coated electrode sheets pass uniformly through multiple ovens at a set speed. The temperature and NMP concentration vary in different ovens. The quality of the electrode sheets leaving the drying process is strongly coupled with the drying process in each oven. Therefore, the battery coating drying system is a typical complex control system characterized by multivariables, nonlinearity, strong coupling, and a certain time delay.

[0003] The battery coating and drying system needs to ensure that the temperature and pressure in each drying chamber remain stable to meet the drying control objectives, while avoiding excessive NMP concentration. Therefore, the design of this control system is a typical multi-objective control problem. However, in current practical operations, most systems employ multiple independent single-loop control designs, with some loops fixed at a setpoint for extended periods. This makes it difficult to adjust in real time according to changes in the actual temperature, air quality, and NMP concentration within the drying chamber. Consequently, in order to prevent NMP from exceeding the limit, the system often operates in overshoot control mode, which, while suppressing the NMP concentration, leads to energy waste.

[0004] Furthermore, different specifications of battery electrode sheets correspond to different dynamic response gains and time delays in the battery coating and drying system, resulting in different amounts of NMP gas evaporation in the oven. Consequently, the temperature and pressure changes in the oven also differ. Different electrode sheets have different optimal operating points. During product switching and standby to production switching processes, there is a lack of real-time controllers. Existing manual operations often involve significantly increasing the circulating air volume to drastically reduce the NMP concentration during switching in order to ensure that the NMP concentration meets the standard. This method of large ventilation and fast circulation leads to excessive low-temperature return air being heated before entering the drying chamber, consuming more electrical energy and causing a large amount of energy waste.

[0005] In summary, the existing automatic control of battery coating and drying systems often uses fixed operating point control, lacking automatic switching of operating points under different operating conditions. This can easily lead to the operating point deviating from the actual needs, resulting in unnecessary energy consumption for heating and wasted energy for fan drive. Summary of the Invention

[0006] To address the problems existing in the prior art, the purpose of this invention is to provide a real-time optimization system for the working point of the drying control system after coating of new energy batteries. This system transforms the dynamic optimization problem of the oven working point in the drying system after battery coating into a linear programming problem with constraint adjustment, thereby achieving the goal of energy-saving and economical operation under the premise that the NMP concentration does not exceed the standard.

[0007] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0008] A real-time optimization system for the working point of the drying control system after coating of new energy batteries. The drying system includes m drying ovens and n sections of main air supply and return ducts. Each drying oven includes an electric return air valve, an internal circulation fan, an electric heating element, and an exhaust fan. The return air volume is controlled by adjusting the frequency of the m internal circulation fans and the opening of the m return air valves; the temperature inside the drying oven is controlled by adjusting the power of the m electric heating elements; the pressure and NMP concentration changes inside the drying oven are compensated by adjusting the frequency of the m exhaust fans; and the NMP concentration of each drying oven section and the total NMP concentration are controlled by adjusting the frequency of the n external circulation fans. Wherein, n is less than m.

[0009] The optimization system uses T RTO To optimize the operating point during the call cycle, the optimization system includes an RTO optimization control module. This module constructs a multi-objective optimization function based on energy consumption and NMP concentration indicators. By solving this objective optimization function, it determines the NMP concentration control target, the internal circulation fan frequency, and the minimum frequency of the external circulation fan. The constructed objective optimization function is implemented using a constrained linear optimization method. Energy consumption is expressed mathematically based on fan frequency and temperature difference, along with the heating output. The NMP concentration indicator is calculated by using a model based on the internal circulation fan frequency to predict the average NMP concentration over a future period. The model for predicting NMP concentration uses either an initial model or a model identified online in real-time during the RTO optimization module call. Each time the RTO optimization control module is called, the optimal NMP concentration control target, internal circulation fan frequency, and lower limit protection values ​​for the external circulation fan frequency are obtained by solving the optimization objective function, thus minimizing the total system airflow while ensuring the NMP concentration does not exceed the limit.

[0010] The objective optimization function of the RTO optimization control module is:

[0011]

[0012] st

[0013] C i <Ci,max i = 1, ..., m (2)

[0014] F in,min ≤f in,i ≤F in,max ,i=1,···,m (3)

[0015] In equation (1), J2 is the objective function, and Q... i (f in,i ), The indicators used to characterize the additional heat and NMP concentration of the oven in section i, respectively; where q i and r i These are the weighting coefficients for the heat index and NMP index of the i-th oven, respectively. The larger the weighting coefficient value, the stricter the constraint on the i-th term in the optimization process; m represents the number of ovens, F in,min and F in,max These are the minimum and maximum frequency constraints for the internal circulation fan, respectively; C i,max This is the maximum safe value for NMP concentration; H c It is a prediction period based on the influence of the internal circulation fan frequency as input on the NMP concentration. The basic time unit for model prediction is the same as the system sampling time. The summation of NMP concentrations and then division by the prediction period time represents the future average NMP concentration corresponding to different internal circulation fan frequency settings in a prediction period.

[0016] Q i It is a representation of the external heat input during the drying process, and the external heat input requires energy consumption, serving as a measure of economic efficiency; Q i and f in The relationship can be expressed by formula (4-6);

[0017] Q i =V i *ΔT i (4)

[0018] ΔT i =T i,tgt -T i,ret (5)

[0019]

[0020] According to equation (4), Q i The internal air circulation volume V of the i-th section of the oven is i and oven temperature increment ΔT i The linear function Q is such that, during normal drying, the return air temperature is lower than the target temperature of the oven, the temperature difference is positive, and the air volume is also non-negative. i The value is positive; where T i,tgtT is the target temperature of the i-th oven. i,ret It is the return air temperature of the i-th oven; where the circulating air volume V i The calculation is performed using formula (6), where f i λ is the fan frequency; p is the number of motor pole pairs; R is the fan rotor radius; L is the fan rotor width; λ v The fan volume coefficient is used; the circulating air volume can be approximated as a linear function of the fan frequency; based on the return air temperature at the RTO optimization time, the internal circulation fan frequency is used. Calculate the heat consumed by the corresponding heating pack [Q1,…,Q] m ];

[0021] C NMP =[C1,…C m [NMP concentration for each section of the oven, C] i NMP concentration is a characterization factor, serving as a variable related to process safety indicators, and is also related to the internal circulation fan frequency f. in Related; considering the dynamic response of NMP and the internal circulation fan frequency, at each dynamic operating point, the NMP concentration [C1,…C] of each oven section is identified. m [Frequency of internal circulation fan] ARX ​​models between:

[0022] A i (z -1 C i (k)=z -d B i (z -1 )f in,i (k)+ξ i (k) i=1,2,…,m (7)

[0023] In the formula, A and B are model parameters, f in,i (k) represents the frequency of the i-th internal circulation fan at time k; C i (k) represents the NMP concentration in the i-th section of the oven at time k; ξ i (k) represents the uncertainty disturbance at time k;

[0024] The internal circulation fan frequency is calculated based on the model in equation (7). The corresponding NMP concentrations [C1,…C1] m The time response of the fan is converted into a continuous time fan frequency and then substituted into the NMP concentration model into the RTO optimization model (1) to solve the optimization problem.

[0025] Before the RTO optimization control module is put into use, the frequency step test of the internal circulation fan is manually set. After the test data is obtained, model identification is performed, and the identified model is stored in the RTO model as the initial model.

[0026] When using the RTO optimization control module for optimization, the initial model is activated in the first cycle, and then each new RTO cycle is judged based on the changes in the internal circulation fan to determine whether the model needs to be updated.

[0027] If, at the k-th RTO optimization time, the absolute value of the change between the optimized internal circulation fan frequency and the original frequency is less than a threshold, the original model continues to be used. If, after RTO optimization, the change in internal circulation fan frequency exceeds the threshold, the model identification data acquisition signal is enabled until the (k+1)-th RTO cycle, at which point model identification data recording ends. The data recording time interval during this period uses the fastest sampling period supported by the system, at least less than 1 / 10 of the RTO cycle, to ensure sufficient sampling points within one cycle. At the (k+1)-th RTO cycle, after optimization output, a model identification is performed to measure the internal circulation fan frequency to NMP concentration. If the accuracy of the identified model is higher than that of the previous model, the model is updated; if the accuracy of the model is lower than that of the previous model, the model is not updated.

[0028] The specific calculation steps for finding the optimal solution to the objective function of the RTO module are as follows:

[0029] Step 1: Construct the Hamiltonian function H(t)

[0030] H(t) = L[x,u,t] + λ T f[x,u,t] (8)

[0031]

[0032] In equation (8), the first term is the performance index function, λ in the second term is the Lagrange operator, and f[x,u,t] is the constraint equation, which is the state equation of the model between the oven NMP concentration and the internal circulation fan frequency in equation (7), where x is the oven NMP concentration and u is the internal circulation fan frequency; according to the principle of minima:

[0033]

[0034]

[0035]

[0036] Since the frequency u of the internal circulation fan is subject to control constraints, it is not necessarily equal to 0, meaning that conditional equation (12) may not have a solution. Considering that if u makes the state vector or Hamiltonian function reach its minimum value, equation (12) can be replaced to minimize the Hamiltonian function. In this case, u is the solution for optimal control, and the corresponding Hamiltonian function is:

[0037] minH[x * ,λ* ,u,t]=H[x * ,λ * ,u * ,t] (13)

[0038] Step 2, given the initial input variable values ​​and learning rate, iteratively solve the problem as follows:

[0039] Given an initial control variable u(0), an initial step size (learning rate) η(0), an iteration cutoff condition ε, and an initial iteration count k = 0, the control parameters are optimized in real time through the following steps (iteration process):

[0040] (1) Calculate the gradient at each step:

[0041] (2) If k = 0, jump to (3); otherwise, substitute u(k) into the objective function J2. If |J2(k) - J2(k+1)| ≤ ε, terminate the iteration and output u(k). If |J2(k) - J2(k+1)| > ε, then calculate Where Δu(k-1)=u(k)-u(k-1),

[0042] (3) Calculation

[0043] (4) Return to (1) and continue to the next iteration;

[0044] The above steps allow for real-time optimization of control parameters, including the internal circulation fan frequency u(t) and the NMP concentration in each oven section.

[0045] Step 3: Based on the optimal internal circulation fan frequency and NMP concentration target value calculated in Step 2, input the optimization objective function into the APC optimization control module;

[0046] Step 4: Based on the optimal internal circulation fan frequency obtained in Step 2, calculate the total duct air volume and the corresponding lower limit protection of the external circulation fan frequency;

[0047] Considering the air volume loss in the return air duct, the total circulating air volume in the return air duct is:

[0048]

[0049] In the formula, V in,i Let V be the circulating air volume of the i-th internal circulation fan. η V represents the total duct airflow loss. out,total This is the sum of the total air volume generated by each internal circulation fan in the return air duct and the total air volume loss in the main duct.

[0050] To ensure that the NMP concentration does not exceed the standard, sufficient return air must enter the oven; that is, the total air volume generated by all external circulation fans must not be less than the total circulating air volume.

[0051]

[0052] Based on the relationship between air volume and fan frequency in equation (6), the reference frequency of the external circulation fan corresponding to the minimum circulating air volume can be calculated, which is the minimum value of the external circulation fan frequency under the current operating conditions, and serves as the lower limit constraint condition for the frequency control of the external circulation fan.

[0053] The model update method for predicting NMP concentration is as follows:

[0054] Based on the optimal internal circulation fan frequency obtained in step 2, check whether the model identification data acquisition signal was enabled in the previous cycle. If the model identification data acquisition signal was enabled in the previous cycle, then perform a model identification from the internal circulation fan frequency to the NMP concentration based on the complete data of one cycle, and at the same time set the model identification acquisition signal to end.

[0055] Based on the optimal internal circulation fan frequency obtained in step 2, compare it with the internal circulation fan frequency value before optimization. If the absolute value is greater than the threshold, enable the model identification data acquisition signal and start recording data for a complete RTO cycle. When the accuracy of the identified model is higher than that of the previous model, update the model for use in the next cycle. If the accuracy of the model is lower than that of the previous model, do not update the model.

[0056] The model for predicting NMP concentration is simplified to an initialization model for different working intervals:

[0057] The main operating frequency range of the internal circulation fan is divided into multiple operating points; the models of multiple operating points are identified offline and saved to the RTO module; during optimization, the corresponding model is used to predict the NMP concentration based on the operating point range of the current internal circulation fan frequency value.

[0058] After adopting the above scheme, this invention constructs a multi-objective optimization function based on energy consumption and NMP concentration indicators. By solving this optimization problem, the control target of NMP concentration and the minimum frequency of the external circulation fan are determined. The objective function of the constructed optimization problem is realized by solving a constrained linear optimization method. Each time the RTO optimization control module is called, the optimal NMP concentration control target and the lower limit protection value of the external circulation fan frequency in the next calling cycle are obtained by solving the optimization problem. This achieves the goal of reducing the total circulating air volume of the system as much as possible without exceeding the NMP concentration limit, thereby achieving the goal of energy-saving and economical operation without exceeding the NMP concentration limit. Attached Figure Description

[0059] Figure 1 A schematic diagram of a battery coating and drying system;

[0060] Figure 2 This is a schematic diagram of the composition of the optimization system of the present invention. Detailed Implementation

[0061] like Figure 1 As shown, the battery coating and drying system includes multiple oven sections, air supply ducts, and return air ducts. Each oven section includes an electric return air valve, an internal circulation fan, an electric heating element, and an exhaust fan. The NMP-containing gas discharged from the multiple oven sections enters the air supply duct through the exhaust ducts, and then the high-temperature, high-NMP-concentration gas is sent to a cooling absorption device. Through heat exchange and cooling, the NMP is liquefied and precipitated. The cooled, low-NMP-concentration circulating air is then blown back into the oven through the return air duct and multiple external circulation fans. During the return air process, a return air valve at the inlet of each oven section controls the return air volume of each section. Simultaneously, to ensure temperature stability within the oven, the circulating return air needs to be preheated through a heat exchanger before entering the oven, continuously replenishing the heat and air consumed by the exhaust air.

[0062] The battery coating and drying system is designed with m drying ovens and n main supply and return air ducts. To ensure temperature and airflow balance within the ovens, the main control loops of the drying system include: adjusting the return air volume by controlling the frequency of m internal circulation fans and the opening of m return air valves; adjusting the oven temperature by controlling the power of m electric heating elements; regulating the NMP concentration by controlling the frequency of n external circulation fans; and compensating for the pressure changes within the oven caused by the actions of other actuators by adjusting the frequency of m exhaust fans, thus maintaining stable operating conditions within the oven. Here, n can be equal to or less than m. In practical applications, several nearby drying ovens often share a single external circulation duct and external circulation fan to reduce costs.

[0063] For each oven section, the controlled variables (CV) include the oven internal temperature, oven internal pressure, and NMP concentration at the oven exhaust outlet. The manipulated variables (MV) include the exhaust fan frequency, return air valve opening, internal circulation fan frequency, electric heating pack power, and external circulation fan frequency. The disturbance variables (DV) include the coating machine speed and the temperature of the returned circulating air after cooling. These variables typically vary depending on the type of battery electrode being dried and the production schedule. Consequently, the oven internal temperature and the minimum external circulation fan frequency will also vary depending on the coating process.

[0064] The main energy consumption in the drying process includes: the motor energy consumption of the exhaust fan and internal circulation fan in each drying chamber, the heating energy of the electric heating pack, and the motor energy consumption of the external circulation fan on the return air duct. The energy balance of the entire process can be characterized by the change in heat per unit time. The electrodes entering the drying chamber are at a relatively low temperature, and the temperature of the electrodes leaving the drying chamber is the final temperature of the drying chamber. At the same time, the airflow entering the exhaust duct loses heat due to the cooling and separation of liquid NMP. Finally, the heating pack raises the temperature to compensate for some of the heat, allowing the temperature inside the drying chamber to reach equilibrium.

[0065] Reference Figure 2 As shown, this invention designs a real-time optimization system for the operating point of a drying control system after coating of new energy batteries, which uses T... RTO The system is updated periodically. The optimization system includes an RTO (Resource-to-Organize) optimization control module. This module constructs a multi-objective optimization function based on energy consumption and NMP concentration indicators. Solving this objective optimization function determines the NMP concentration control target, the internal circulation fan frequency, and the minimum frequency of the external circulation fan. The constructed objective optimization function is implemented using a constrained linear optimization method. Energy consumption is expressed mathematically based on fan frequency and temperature difference, along with heating output. The NMP concentration indicator is calculated by predicting the average NMP concentration over a future period using a model based on the internal circulation fan's NMP concentration. The model for predicting NMP concentration uses either an initial model or a model identified online in real-time as needed during the RTO optimization module call. Each time the optimization system is called, the optimal NMP control target and the lower limit protection value of the external circulation fan frequency for the next call cycle are obtained by solving the optimization problem. This aims to minimize the total circulating air volume while keeping the NMP concentration within limits. Since the total circulating air volume and total system power consumption are positively correlated, the optimization objective reflects the goal of reducing energy consumption.

[0066] Before the RTO optimization control module is put into use, the frequency step test of the internal circulation fan is manually set. After the test data is obtained, model identification is performed, and the identified model is stored in the RTO model as the initial model.

[0067] When using the RTO optimization control module for optimization, the initial model is activated in the first cycle, and then each new RTO cycle is judged based on the changes in the internal circulation fan to determine whether the model needs to be updated.

[0068] If, at the k-th RTO optimization time, the absolute value of the change between the optimized internal circulation fan frequency and the original frequency is less than a threshold, the original model continues to be used. If, after RTO optimization, the change in internal circulation fan frequency exceeds the threshold, the model identification data acquisition signal is enabled until the (k+1)-th RTO cycle, at which point model identification data recording ends. The data recording time interval during this period uses the fastest sampling period supported by the system, at least less than 1 / 10 of the RTO cycle, to ensure sufficient sampling points within one cycle. At the (k+1)-th RTO cycle, after optimization output, a model identification is performed to measure the internal circulation fan frequency to NMP concentration. If the accuracy of the identified model is higher than that of the previous model, the model is updated; if the accuracy of the model is lower than that of the previous model, the model is not updated.

[0069] The optimization system of this invention transforms the dynamic optimization problem of the working point of the battery coating oven into a linear programming problem with constraint adjustment, thereby achieving the goal of energy-saving and economical operation under the premise that the NMP concentration does not exceed the standard.

[0070] Considering the process characteristics and energy balance of electrically heated hot air circulating drying, a multi-objective optimization function J2 based on energy consumption and NMP concentration indices is determined. Finally, a constrained linear optimization problem is solved to minimize the total circulating air volume of the system while keeping the NMP concentration within limits. Taking an m-section drying oven as an example, the internal circulation fan frequency is... The optimization model for the RTO optimization control module in the battery coating and drying process is as follows:

[0071]

[0072] st

[0073] C i <C i,max i = 1, ..., m (2)

[0074] F in,min ≤f in ≤F in,max (3)

[0075] In the formula, J2 is the objective function, and Q... i (f in,i ), The indicators used to characterize the additional heat and NMP concentration of the oven in section i, respectively; where q i and r i These are the weighting coefficients for the heat index and the NMP index, respectively. The larger the weighting coefficient value, the stricter the constraint on the corresponding i-th term in the optimization process; m represents the number of ovens, F in,min and F in,max These are the minimum and maximum frequency constraints for the internal circulation fan, respectively. C i,maxThis represents the maximum safe value for NMP concentration. Hc is the prediction period based on the influence of the internal circulation fan frequency as input on NMP concentration; the length of the period is determined by the model response time. The summation of NMP concentrations divided by the prediction time represents the future average NMP concentration for different internal circulation fan frequency settings within a prediction period.

[0076] Q i It is a representation of the external heat input during the drying process, and since external heat input requires energy consumption, it can be used as a measure of economic efficiency; Q i and f in The relationship can be expressed by formula (4-6);

[0077] Q i =V i *ΔT i (4)

[0078] ΔT i =T i,tgt -T i,ret (5)

[0079]

[0080] According to equation (4), Q i The internal air circulation volume V of the i-th section of the oven is i and oven temperature increment ΔT i The linear function Q is such that, during normal drying, the return air temperature is lower than the target temperature of the oven, the temperature difference is positive, and the air volume is also non-negative. i The value is positive. Where T... i,tgt T is the target temperature of the i-th oven. i,ret It is the return air temperature of the i-th oven; where the circulating air volume V i It can be calculated using formula (6), where f i Fan frequency; p is the number of motor pole pairs; R is the rotor radius; L is the rotor width; λ v The fan volume coefficient is used; the recirculated air volume can be approximated as a linear function of the fan frequency. Based on the return air temperature at the RTO optimization time, the internal recirculation fan frequency can be used as a parameter. Calculate the heat consumed by the corresponding heating pack [Q1,…,Q] m ].

[0081] C NMP =[C1,…C m [NMP concentration for each section of the oven, C] i It is a characterization of NMP concentration, and as a variable related to process safety performance, it is also related to the internal circulation fan frequency f. inRelated. Considering the dynamic response of NMP and the internal circulation fan frequency, the NMP concentration [C1,…C] of each section of the oven was identified at each dynamic operating point. m [Frequency of internal circulation fan] ARX ​​models between:

[0082] A i (z -1 C i (k)=z -d B i (z -1 )f in,i (k)+ξ i (k) i=1,2,…,m (7)

[0083] In the formula, A and B are model parameters, f in,i (k) represents the frequency of the i-th internal circulation fan at time k; C i (k) represents the NMP concentration in the i-th section of the oven at time k; ξ i (k) represents the uncertainty disturbance at time k;

[0084] The frequency of the internal circulation fan can be calculated according to equation (7). The corresponding NMP concentrations [C1,…C1] m The time response is converted into a continuous-time model and substituted into the RTO optimization model (1) to solve the optimization problem. According to the RTO optimization model (1), as the frequency of the internal circulation fan decreases, the air volume of the oven decreases, which reduces the heat exchange between the circulating air and the total heat consumed by each heating pack in the oven decreases, i.e., the first term decreases; at the same time, the total NMP concentration in each section of the oven increases due to the decrease in return air volume, i.e., the second term increases. Therefore, due to the mutual cancellation between the decrease in total heat consumption and the increase in total NMP concentration, the target frequency F of the internal circulation fan at the optimal operating point can be calculated. in,tgt And the optimal NMP concentration control target for each section is calculated using model (7). The optimal control objective of CV under the current operating conditions is y TGT Furthermore, rolling optimization is performed at each sampling time, ultimately only the first control input is applied to the system.

[0085] The specific calculation steps for finding the optimal solution to the objective function of the RTO optimization control module are as follows:

[0086] Step 1: Construct the Hamiltonian function H(t)

[0087] H(t) = L[x,u,t] + λ T f[x,u,t] (8)

[0088]

[0089] In equation (8), the first term is the performance index function, λ in the second term is the Lagrange operator, and f[x,u,t] is the constraint equation, which is the state equation of the model between the oven NMP concentration and the internal circulation fan frequency in equation (7), where x is the oven NMP concentration and u is the internal circulation fan frequency; according to the principle of minima:

[0090]

[0091]

[0092]

[0093] Since the frequency u of the internal circulation fan is subject to control constraints, it is not necessarily equal to 0, meaning that conditional equation (12) may not have a solution. Considering that if u makes the state vector or Hamiltonian function reach its minimum value, equation (12) can be replaced to minimize the Hamiltonian function. In this case, u is the solution for optimal control, and the corresponding Hamiltonian function is:

[0094]

[0095] Step 2: Given the initial input variable values ​​and step size, iteratively solve the problem as follows.

[0096] Given an initial control variable u(0), an initial step size (learning rate) η(0), an iteration cutoff condition ε, and an initial iteration count k = 0, the control parameters are optimized in real time through the following steps (iteration process):

[0097] (1) Calculate the gradient at each step:

[0098] (2) If k = 0, jump to (3); otherwise, substitute u(k) into the objective function J2. If |J(k) - J(k+1)| ≤ ε, terminate the iteration and output u(k). If |J(k) - J(k+1)| > ε, then calculate Where Δu(k-1)=u(k)-u(k-1),

[0099] (3) Calculation

[0100] (4) Return to (1) and continue to the next iteration;

[0101] The above steps allow for real-time optimization of control parameters, including the internal circulation fan frequency u(t) and the NMP concentration in each oven section.

[0102] Step 3: Based on the optimal internal circulation fan frequency and NMP concentration target value calculated in Step 2, output the target value to other controllers or operators as the control target value.

[0103] Step 4: Based on the optimal internal circulation fan frequency obtained in Step 2, calculate the total duct air volume and the corresponding lower limit protection of the external circulation fan frequency.

[0104] Considering the air volume loss in the return air duct, the total circulating air volume in the return air duct is:

[0105]

[0106] In the formula, V in,i Let V be the circulating air volume of the i-th internal circulation fan. η V represents the total duct airflow loss. out,total This is the sum of the total air volume generated by each internal circulation fan in the return air duct and the total air volume loss in the main duct.

[0107] To ensure that the NMP concentration does not exceed the standard, sufficient return air must enter the oven; that is, the total air volume generated by all external circulation fans must not be less than the total circulating air volume.

[0108]

[0109] Based on the relationship between air volume and fan frequency in equation (6), the reference frequency of the external circulation fan corresponding to the minimum circulating air volume can be calculated, which is the minimum value of the external circulation fan frequency under the current operating conditions, and serves as the lower limit constraint condition for the frequency control of the external circulation fan.

[0110] Step 5: Based on the optimal internal circulation fan frequency obtained in Step 2, check whether the model identification data acquisition signal was enabled in the previous cycle. If the model identification data acquisition signal was enabled in the previous cycle, then perform a model identification on the internal circulation fan frequency and NMP concentration based on the complete data of one cycle, and at the same time set the model identification acquisition signal to end.

[0111] Step 6: Based on the optimal internal circulation fan frequency obtained in Step 2, compare it with the internal circulation fan frequency value before optimization. If the change is significant (absolute value greater than a threshold, such as 2Hz), enable the model identification data acquisition signal and start recording data for a complete RTO cycle. The data recording cycle is based on the actual acquisition cycle and needs to be at least 1 / 10 or less of the RTO cycle (in practice, the sampling cycle is 5 seconds and the RTO cycle is 3 minutes, which is 60 times the sampling cycle). When the accuracy of the identified model is higher than that of the previous model, update the model for the next cycle; if the model accuracy is lower than that of the previous model, do not update the model.

[0112] The model used for predicting NMP concentration during RTO optimization can employ either the real-time update strategy described in steps 5 and 6, or it can be simplified to automatic selection of the initial model within different operating ranges. For example, the main operating frequency range of the internal circulation fan [18-46] Hz can be divided into 7 operating points, each covering a 4 Hz range. The models for the 7 operating points are identified offline and saved to the RTO module. During optimization, the corresponding model is activated to predict NMP concentration based on the operating point range of the current internal circulation fan frequency value.

[0113] In practical applications, the real-time optimization system of this invention can dynamically output control target values ​​for different operating points of the drying system based on a comprehensive index of energy consumption and NMP concentration. The output control target values ​​can provide guidance for operators to manually adjust, allowing operators to decide whether to accept the output, or can be automatically output to a traditional single-loop PID controller as a control target setpoint to achieve energy saving under the dynamic target of the system. It can also be combined with other multivariable controllers to form a two-layer optimization control, further improving control accuracy and energy saving.

[0114] The above description is merely an embodiment of the present invention and does not constitute any limitation on the technical scope of the present invention. Therefore, any minor modifications, equivalent changes, and alterations made to the above embodiments based on the technical essence of the present invention shall still fall within the scope of the technical solution of the present invention.

Claims

1. A real-time optimization system for working points of a new energy battery post-coating drying control system, the drying control system comprising m a drying oven and a matching n-section total air supply pipe and total return air pipe, wherein each section of the drying oven comprises an electric return air valve, an internal circulation air fan, an electric heating pack, and an exhaust fan; and characterized in that: By adjusting m the frequency of the inner circulating fan and m the opening of the return air valve to control the return air volume; by adjusting m the power of the electric heating package to control the temperature in the oven; by adjusting m the frequency of the exhaust fan to compensate for the change of pressure in the oven and the change of NMP concentration; by adjusting n the frequency of the outer circulating fan to control the NMP concentration in each section of the oven and the total NMP concentration; wherein, n less than m ; The optimization system is configured to T RTO To achieve the working point optimization for the calling period, the optimization system comprises an RTO optimization control module; the RTO optimization control module constructs a multi-objective optimization function based on energy consumption and NMP concentration indicators, determines the control target of NMP concentration, the frequency of the inner circulating fan and the minimum frequency of the outer circulating fan by solving the objective optimization function, and the constructed objective optimization function is achieved by solving a linear optimization method with constraints; wherein the indicator of energy consumption is the newly added heat established by the fan frequency and the temperature difference; the NMP concentration indicator is the average NMP concentration in the future period of time calculated by the model from the inner circulating fan to the NMP concentration; the model for predicting the NMP concentration uses the model given by initialization or the model identified in real time online according to the needs during the calling of the RTO optimization control module; the optimal NMP concentration control target, the frequency of the inner circulating fan and the lower limit protection value of the frequency of the outer circulating fan in the next calling period are obtained by solving the optimization objective function each time the RTO optimization control module is called, so as to reduce the total circulating air volume of the system as much as possible under the premise that the NMP concentration does not exceed the standard. 2.The new energy battery post-coating drying control system working point real-time optimization system according to claim 1, characterized in that: The target optimization function of the RTO optimization control module is: （1） s.t. （2） （3） In formula (1), J 2 is the target function, , respectively represent the first i section oven added heat and NMP concentration indicators; wherein and are the heat indicators and NMP indicators of the first i section oven weight coefficients, the greater the weight coefficient value, the more stringent the optimization process corresponding i term constraint; m represents the number of ovens. and These are the minimum and maximum value constraints for the internal circulation fan frequency, respectively. This is the maximum safe value for NMP concentration; H c It is a prediction period based on the influence of the internal circulation fan frequency as input on the NMP concentration. The basic time unit for model prediction is the same as the system sampling time. The summation of NMP concentrations and then division by the prediction period time represents the future average NMP concentration corresponding to different internal circulation fan frequency settings in a prediction period. is a representation of the external input heat during the drying process, and the external heat input requires energy consumption, which is a measure of economic indicators; and The relationship between and can be expressed by equations (4)-(6); （4） （5） （6） According to formula (4), is the first i section oven target temperature, is the first section oven air volume, is the first section oven temperature increment, i is the first section oven return air temperature, i is the first section oven air volume, is the fan frequency, p is the number of motor magnetic poles, L is the fan blade rotor radius, is the fan blade rotor width, is the fan volume coefficient, is the heat consumption of the corresponding heating bag. NMP concentration in each zone oven, is the NMP concentration characterization, related to the process safety indicator variable, also related to the inner loop fan frequency ; considering the dynamic response of NMP and inner loop fan frequency, at each dynamic operating point, the ARX model between NMP concentration in each zone oven and inner loop fan frequency is identified: （7） In the formula, , For model parameters, for k Time of the first i The frequency of the internal circulation fan; for k Time of the first i NMP concentration in the oven; for k Momentary uncertainty disturbance; The inner recycle blower frequency is calculated from the model of equation (7) The corresponding NMP concentration The time response is converted to continuous time blower frequency versus NMP concentration model and inserted into the objective optimization function (1) of the RTO optimization control module to solve the optimization problem. 3.The working point real-time optimization system of a new energy battery post-coating drying control system according to claim 2, characterized in that: Before the RTO optimization control module is put into operation, the frequency step test of the inner circulating fan is manually set, the test data is obtained, and then model identification is performed, and the identified model is stored in the target optimization function of the RTO optimization control module as an initial model for saving; When the RTO optimization control module is used for optimization, the initial model is enabled in the first cycle, and then whether the model needs to be updated is judged according to the change of the inner circulating fan in each new RTO optimization cycle; If the absolute value of the change of the inner circulating fan frequency after RTO optimization and before optimization is less than a threshold value at the kth RTO optimization cycle, the original model is continuously used; if the change of the inner circulating fan frequency after RTO optimization exceeds the threshold value, the model identification data collection signal is enabled, and the model identification data recording is ended at the k+1th RTO optimization cycle; the data recording time interval during the period is the fastest sampling period supported by the system, which is at least 1 / 10 of the RTO optimization cycle, to ensure that there are enough sampling points in one cycle; At the k+1th RTO optimization cycle, after optimization output, a model identification of the inner circulating fan frequency to the NMP concentration is performed; when the precision of the identified model is higher than that of the previous model, the model is updated; if the model precision is lower than that of the previous model, the model is not updated. 4.The working point real-time optimization system of a new energy battery post-coating drying control system according to claim 3, characterized in that: The specific calculation steps of the optimization target function optimal solution problem of the RTO optimization control module are as follows: Step 1, constructing the Hamiltonian function （8） （9） In formula (8), the first term is a performance index function, and the second term is a Lagrange operator, is a Lagrange operator, is a constraint equation, i.e., a state equation of the model between the oven NMP concentration and the inner circulating fan frequency in formula (7), is the oven NMP concentration, is the inner circulating fan frequency; according to the minimum value principle: （10） （11） （12） Due to the inner loop blower frequency There are control constraints, not necessarily equal to 0, that is, the condition equation (12) is not necessarily a solution; consider if Make the state vector or Hamilton function minimum, replace equation (12) to make the Hamilton function minimum, and The optimal control solution is the minimum value of the Hamilton function, and the corresponding Hamilton function is: （13） Step 2, given the initial input variable value and learning rate, the optimal solution is iteratively solved as follows: given initial control variable , initial step size, i.e. learning rate , iteration stopping condition , initial iteration count Real-time optimization of the control parameters is accomplished by the following steps: (1) Calculate the gradient of each step: ; (2) If , go to (3); otherwise, substitute into the objective function , if , terminate the iteration and output , if , compute , where , ; (3) calculating ; (4) return (1) to continue the next iteration; By the above steps, the inner circulating fan frequency can be optimized in real time and NMP concentration index of each section of oven ; Step 3, based on the optimal inner circulating fan frequency and NMP concentration target value calculated in step 2, the control target value is output to other controllers or operators; Step 4, based on the optimal inner circulating fan frequency obtained in step 2, the total pipe air volume and the corresponding outer circulating fan frequency lower limit protection are calculated; Considering the air volume loss of the return air pipe, the total circulating air volume in the return air pipe is: （14） wherein is the total air volume loss of the duct, i is the circulating air volume of the i-th inner circulating fan, is the total air volume loss of the duct, is the sum of the total air volume generated by the inner circulating fans in the return air duct and the total air volume loss of the duct. In order to ensure that the NMP concentration does not exceed the standard, sufficient return air must enter the oven, that is, the total air volume generated by each outer circulating fan cannot be less than the total circulating air volume: （15） According to the relationship between the air volume and the fan frequency in formula (6), the minimum circulating air volume corresponding to the reference frequency of the outer circulating fan can be calculated, that is, the minimum value of the outer circulating fan frequency under the current working condition is used as the lower limit constraint condition of the outer circulating fan frequency control. 5.The working point real-time optimization system of a new energy battery post-coating drying control system according to claim 4, characterized in that: The model updating method for predicting the NMP concentration is as follows: Based on the optimal inner circulating fan frequency obtained in step 2, it is checked whether the model identification data collection signal is enabled in the last cycle, if the model identification data collection signal is enabled in the last cycle, a model identification of the inner circulating fan frequency to the NMP concentration is performed according to the complete data of one cycle, and the model identification collection signal is set to end; Based on the optimal inner circulating fan frequency obtained in step 2, and the inner circulating fan frequency value before optimization, if the absolute value is greater than a threshold, a model identification data acquisition signal is set to be enabled, and data of a complete RTO optimization cycle is recorded; when the accuracy of the identified model is higher than that of the previous model, the model is updated for use in the next cycle; if the model accuracy is lower than that of the previous model, the model is not updated. 6.The working point real-time optimization system of a new energy battery post-coating drying control system according to claim 4, characterized in that: The model for predicting the NMP concentration is simplified into initialization models in different working intervals: The main operating frequency range of the inner circulating fan is divided into multiple working points; models of the multiple working points are identified offline and saved to an RTO optimization control module; during optimization, according to the working point range in which the current inner circulating fan frequency value is located, a corresponding model is enabled to predict the NMP concentration.

Images