Thermal management control method for MPC grinding machine based on finite element

Abstract

The invention provides an MPC grinding machine thermal management control method based on a finite element, and the method comprises the steps: building a physical model related to the thermal management of a grinding machine system, and taking finite element simulation data as a prediction sample space; according to the temperature delay of the grinding machine system, a target transfer function controlled by a prediction model is established to serve as a control equation of model prediction; performing rolling optimization and feedback correction on the prediction model through an MPC control algorithm, and fusing the predicted temperature change trend with real data measured by a temperature sensor to obtain a fused optimal output variable quantity; and the input flow is determined according to the optimal output variable quantity, so that the real-time temperature control effect of the grinding machine system is achieved. While the control effect is improved, predictive control under different working condition requirements can be achieved, combined control over multiple grinding machines is achieved, and more reasonable heat management control is achieved by combining the particle swarm optimization algorithm with coupling adjustment of the Gaussian process.

Description

Technical Field

[0001] The present invention relates to the technical field of thermal management of powder materials and liquid slurry micro-nano grinders, and particularly relates to a thermal management control method for an MPC grinder based on finite element. Background Art

[0002] The development of new material technologies has not only promoted the progress of powder material micro-nano grinding technologies and control systems, but also put forward higher requirements for the particle size and production capacity of ground powder materials. During the process of grinding powder materials, the control of particle size, material contamination, and production capacity improvement are key factors. The control of particle size directly affects the performance and application fields of materials. Currently, through physical and chemical methods, wet grinding of solid particles in a liquid medium can achieve a fineness of less than 100 nanometers. The improvement of production capacity and industrial production are related to production costs and efficiency. Among them, the heat generation and temperature control during the grinding process will greatly affect the particle size and production capacity of new materials. Traditional grinders use a water-cooling method, and most use algorithms such as PI control in the control method. However, there is no accurate modeling analysis for the heat generation effect during the grinding process, resulting in incomplete estimation of the heat generation amount, or directly using the on-line power as a reference value for the heat generation amount. Moreover, in the cross-condition joint control of multiple devices, due to the large instantaneous cooling water consumption, the equipment selection is too large, the upfront investment is too large, and the production capacity is unstable, etc. Considering that the on-line power of the equipment itself does not fully represent the heat generation amount, the machine cannot be equivalent to a heat-generating body. The equipment only generates heat during the grinding process when the rotor drives the grinding medium to shear and rub against the slurry. A considerable part of the power is used in the grinding and pulverization process. Therefore, using the on-line power as a reference benchmark, the control effect is not perfect, and the stabilization time and overshoot will be relatively large. Summary of the Invention

[0003] The object of the present invention is: aiming at the deficiencies existing in the above background art, to provide an improved solution, introducing the finite element simulation results as the reference value of the heat generation amount for a specific period of time, and at the same time introducing the MPC control algorithm, and through the rolling optimization method, effectively tracking the optimal water-saving path to achieve optimal control.

[0004] To achieve the above object, the present invention provides a thermal management control method for an MPC grinder based on finite element, including the following steps: S1, establishing a physical model related to the thermal management of the grinder system, and using the finite element simulation data as the prediction sample space; S2, according to the delay of the grinder system temperature, establishing the target transfer function of the predictive model control as the control equation of the model prediction; S3. Through the MPC control algorithm, perform rolling optimization and feedback correction on the prediction model, and fuse the predicted temperature change trend with the real data measured by the temperature sensor to obtain the optimal output change amount after fusion. Among them, MPC (Model Predictive Control) is model predictive control, aiming to generate optimal control inputs by predicting future system states and considering constraints; S4. Determine the input flow rate according to the optimal output change amount to achieve the effect of real-time temperature control of the grinder system.

[0005] Further, the physical model established in S1 is: ; Among them, is the specific heat capacity of the grinding slurry, with the unit of ; is the specific heat capacity of the cooling water, with the unit of ; is the outlet temperature of the grinding slurry, with the unit of ; is the outlet temperature of the cooling water, with the unit of ; is the density of the grinding slurry, with the unit of ; is the density of the cooling water, with the unit of ; is the volume inside the grinder, with the unit of ; is the volume of the cooling jacket, with the unit of ; is the input flow rate of the grinding slurry, with the unit of ; is the input flow rate of the cooling water, with the unit of ; is the inlet temperature of the grinding slurry, with the unit of ; is the inlet temperature of the cooling water, with the unit of ; W is the heat generation of the grinder, with the unit of ; U is the heat transfer coefficient, with the unit of ; S is the heat transfer area, with the unit of .

[0006] Further, the heat generation W of the grinder in S1 uses the result of finite element simulation as the input value, sets the input flow rate of the grinding slurry as a fixed value, sets the input flow rate of the cooling water as the input value, and sets the outlet temperature of the grinding slurry and the outlet temperature of the cooling water as the output values.

[0007] Further, the target transfer function in S2 is: ; Wherein, is the output matrix of the system, including the outlet temperature of the grinding slurry and the outlet temperature of the cooling water ; is the input value of the system, i.e., the input flow rate of the cooling water ; is the input change value of the system; is the output weight matrix of the system, , is the weight factor of the cooling water temperature, is the weight factor of the grinding slurry temperature; is the weight factor of the cooling water flow rate.

[0008] Further, according to the requirements of the actual project, different weight factors are used to adjust the weights between the cooling water and the grinding slurry. When the value of increases and the value of decreases, it focuses on controlling the temperature. When the value of decreases and the value of increases, temperature control is carried out on the premise of saving flow rate.

[0009] Further, S3 specifically includes the following sub-steps: S31, in each sampling period, optimize the inlet temperature and outlet temperature of the grinding machine system at adjacent times, find the optimal control sequence of the predicted value to predict the outlet temperature value of the grinding machine system at the next moment, obtain the controlled variable of the controlled object according to the predicted value of the outlet temperature at the next moment, and perform the next prediction initialization, and continue to iterate and advance following the optimization time; S32, during each sampling period, compensate for the temperature trend error generated by the prediction model by collecting the measured outlet temperature value of the grinding machine in real time.

[0010] Further, for the joint control of multiple grinding machines, set 2s devices, divided into two working conditions: rough grinding with s devices and fine grinding with s devices. During the trial production process, adjust the two working conditions respectively, and collect the adjusted T and ω, where is the start-up time of the device, is the production cycle, the value of and the value of represent the rough grinding working condition, the value of and Value represents the fine grinding condition, represents the cooling water input of the i-th equipment during coarse grinding, represents the cooling water input of the i-th equipment during fine grinding, T, ω, , all follow the Gaussian distribution , , , , and ; The PSO algorithm is used for optimization. The PSO (Particle Swarm Optimization) algorithm is the particle swarm optimization algorithm. Set the total water consumption as G, the cooling water input of each equipment as Q, and the value of the coarse grinding condition Value , the fine grinding condition Value , discretize the function G. The sampling discretization is based on the Gaussian distribution function, so that the sampling density is high in the early stage of equipment startup and decreases after stabilization; after determining the sampling point positions, use the linear interpolation method to select the and corresponding to M sampling points, convolve with the square wave function to form the function G, and set the time difference of the square wave function based on the different water demand in different time periods; optimize with the PSO algorithm to obtain the optimized T value, and record the actual production situation at the same time. Record the actual production cycle and start / stop time of each equipment, return the earliest updated T value and ω value, and obtain the jointly controlled T value after several iterations.

[0011] Furthermore, the sampling formula of the Gaussian distribution function is: ; where m represents the sampling point, and M sampling points are set, ; is the corresponding time point; are all fitting parameters, obtained during the actual project debugging, w 1 represents the peak height of the Gaussian function, that is, the maximum value of the function at the center position, w 2 corresponds to the mean value of the Gaussian distribution, that is, the position of the distribution symmetry axis, w 3 is related to the variance.

[0012] Furthermore, the function G is specifically: .

[0013] Furthermore, for the cross-condition joint control of multiple grinding machines, when statistically analyzing the Gaussian distribution, data fusion is performed on the T values and ω values of each device. is the startup time of the device. is the production cycle and follows a Gaussian distribution. 、 , is the cooling water input of the i-th device during rough grinding. is the cooling water input of the i-th device during fine grinding. After fusion, it follows a Gaussian distribution. ; The sampling formula of the Gaussian distribution function is: ; where m represents the sampling point, and M sampling points are set. ; is the corresponding time point. are all fitting parameters obtained during the actual project debugging. w 1 represents the peak height of the Gaussian function, that is, the maximum value of the function at the center position. w 2 corresponds to the mean value of the Gaussian distribution, that is, the position of the distribution symmetry axis. w 3 is related to the variance. Using the PSO algorithm combined with the Gaussian process, set the total water consumption as G: ; Optimize the G function using the PSO algorithm to obtain the optimized T value and ω value. After several iterations, the T value for joint control is obtained.

[0014] The above solution of the present invention has the following beneficial effects: The thermal management control method of the MPC grinding machine based on the finite element provided by the present invention, based on the grinding machine system, establishes a physical model related to thermal management, and uses the result of finite element simulation as the heat generation reference value for a specific period of time. At the same time, the MPC control algorithm is introduced, and through the method of rolling optimization, it effectively tracks the optimal water-saving path. While improving the control effect, the controller can use different weight factors according to the requirements of the actual project to adjust the weights between the cooling water and the grinding slurry, realizing predictive control under different working conditions requirements, and achieving the purpose of energy conservation and water consumption reduction. The present invention is directed to the joint control of multiple grinding machines. Through the coupled adjustment of the particle swarm optimization algorithm combined with the Gaussian process, a reasonable control objective can be achieved. It can not only adjust the production cycle of each device, stagger the stages with high heat generation during feeding and the initial grinding stage, but also have less impact on the total amount of cooling water, avoiding the problem of uneven distribution of cooling water in the initial operation stage. This can not only reduce production interruptions and reject rates caused by temperature fluctuations, but also improve production efficiency, product quality stability, and reduce production costs. Other beneficial effects of the present invention will be described in detail in the subsequent specific implementation section. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 is the step flow chart of the present invention; Figure 2 is the schematic diagram of the grinding machine system of the present invention; Figure 3 is the joint control flow chart of multiple grinding machines in two types of working conditions of the present invention; Figure 4 is the joint control flow chart of multiple grinding machines across working conditions of the present invention. SPECIFIC IMPLEMENTATION MODE

[0016] The following specific examples illustrate the implementation modes of the present disclosure. Those skilled in the art can easily understand other advantages and effects of the present disclosure from the content disclosed in this specification. Obviously, the described embodiments are only part of the embodiments of the present disclosure, not all of them. The present disclosure can also be implemented or applied through other different specific implementation modes. Various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present disclosure. It should be noted that, without conflict, the following embodiments and the features in the embodiments can be combined with each other. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present disclosure without creative efforts belong to the scope of protection of the present disclosure.

[0017] It should be noted that the following describes various aspects of embodiments within the scope of the appended claims. It should be apparent that the aspects described herein can be embodied in a wide variety of forms, and any specific structure and / or function described herein is illustrative only. Based on the present disclosure, those skilled in the art should understand that one aspect described herein can be implemented independently of any other aspect, and two or more of these aspects can be combined in various ways. For example, any number of aspects described herein can be used to implement the device and / or practice the method. Additionally, this device and / or this method can be implemented using other structures and / or functions in addition to one or more of the aspects described herein.

[0018] It should also be noted that the diagrams provided in the following embodiments only illustrate the basic concept of the present disclosure schematically. The diagrams only show the components related to the present disclosure and are not drawn according to the number, shape, and size of the components in actual implementation. The types, quantities, and ratios of the components in actual implementation can be arbitrarily changed, and the component layout type may also be more complex. Additionally, in the following description, specific details are provided to facilitate a thorough understanding of the examples. However, those skilled in the art will understand that the aspects described can be practiced without these specific details.

[0019] As Figure 1 shown, an embodiment of the present invention provides a finite element-based thermal management control method for an MPC grinder. The grinder system involved is as Figure 2 shown, and the method may specifically include the following steps: S1, establish a physical model related to the thermal management of the grinder system, and use the finite element simulation data as the prediction sample space.

[0020] In this step, the established physical model is specifically: ; wherein, is the specific heat capacity of the grinding slurry, with the unit of ; is the specific heat capacity of the cooling water, with the unit of ; is the outlet temperature of the grinding slurry, with the unit of ; is the outlet temperature of the cooling water, with the unit of ; is the density of the grinding slurry, with the unit of ; is the density of the cooling water, with the unit of ; is the volume inside the grinder, with the unit of ; is the volume of the cooling jacket, with the unit of ; is the input flow rate of the grinding slurry, with the unit of ; is the input flow rate of the cooling water, with the unit of ; is the inlet temperature of the grinding slurry, with the unit of ; is the inlet temperature of the cooling water, with the unit of ; W is the heat generation of the grinder, with the unit of ; U is the heat transfer coefficient, with the unit of ; S is the heat transfer area, with the unit of .

[0021] In this step, the value of the heat generated by the grinder itself, W, is used as the input value based on the results of finite element simulation. At the same time, based on the actual situation, the input flow rate of the grinding slurry is set to a fixed value, and only the temperature of the grinder system is controlled by controlling the flow rate of the cooling water. Therefore, the input flow rate of the cooling water can be set as the input value, and the outlet temperature of the grinding slurry and the outlet temperature of the cooling water are set as the output values. Since it is a single-input multi-output system, the control stability will deteriorate. Therefore, a controller for the prediction model needs to be designed to stabilize the system through model prediction.

[0022] S2. According to the delay of the temperature of the grinder system, establish the target transfer function of the predictive model control as the control equation of the model prediction.

[0023] In this step, the target transfer function is specifically: ; where is the output matrix of the system, including the outlet temperature of the grinding slurry and the outlet temperature of the cooling water ; is the input value of the system, which is the input flow rate of the cooling water , is the input change value of the system; is the output weight matrix of the system, , is the weight factor of the cooling water temperature, is the weight factor of the grinding slurry temperature; is the weight factor of the cooling water flow rate.

[0024] Therefore, the controller can use different weight factors according to the requirements of the actual project to adjust the weights between the cooling water and the grinding slurry. If has a large value and has a small value, it means that the project itself focuses on controlling the temperature and has few restrictions on the flow rate. On the contrary, the temperature is controlled from the perspective of saving the flow rate. Thus, by setting different weight factors, the predictive control under different working conditions of the model can be realized. For example, the weight factor of the outlet temperature of the cooling water can be increased to affect the optimal value of the controller, so as to better control the temperature. The optimal control of the cooling water flow rate can also be adjusted by adjusting the value of to adapt to the control under different working conditions and achieve the purpose of energy saving and reducing water consumption.

[0025] S3. Through MPC control, perform rolling optimization and feedback correction on the prediction model, and fuse the predicted temperature change trend with the real data measured by the temperature sensor to obtain the optimal output change amount after fusion.

[0026] This step can specifically include the following two sub-steps: S31. Rolling optimization of the prediction model: In each sampling period, optimize the inlet temperature and outlet temperature of the grinding machine system at adjacent moments, find the optimal control sequence of the predicted value to predict the outlet temperature value of the grinding machine system at the next moment, obtain the controlled variable of the controlled object according to the predicted value of the outlet temperature at the next moment, and perform the next prediction initialization, and continue to iterate and advance following the optimization time in this way; In this step, discretize the prediction model: ; Among them, is the sampling time; is the state vector; is the input; contains the disturbance term, such as the heat generation W of the grinding machine, the inlet temperature of the grinding slurry and the inlet temperature of the cooling water ; is the matrix after discretization.

[0027] At each sampling time , MPC solves the optimization problem for the next N steps, and the objective function is: ; Among them, is the set temperature of the outlet of the grinding slurry and the set temperature of the outlet of the cooling water; the weight matrix , , are the weight factors corresponding to the temperature of the grinding slurry, the temperature and flow rate of the cooling water respectively; the predicted output is calculated based on the current state and the future input sequence .

[0028] When performing rolling optimization, use the discrete model to predict the output for the next N steps: ; Solve the optimal input sequence through quadratic programming (QP); only apply the optimal input at the current moment , discard the subsequent inputs; at the next moment , repeat the above process.

[0029] S32, Feedback correction of the prediction model: During each sampling period, by collecting the measured outlet temperature of the grinder in real time, a certain compensation is made for the temperature trend error generated by the prediction model.

[0030] Specifically, based on the error compensation mechanism, at the end of each sampling period, the prediction model is corrected through the actually measured outlet temperature including: Calculating the prediction error: ; where is the predicted value of the current output at the previous moment; Updating the state estimation: ; where L is the correction gain matrix, such as the Kalman filter gain.

[0031] Disturbance estimation: If there are unmodeled disturbances (such as W fluctuations), they are modeled as additional states and updated online.

[0032] Through the feedback correction of this step, the system can eliminate the cumulative error caused by model mismatch, such as the deviation of the heat transfer coefficient U, adapt to external disturbances, and improve robustness.

[0033] S4, Determine the input flow rate according to the optimal output change amount to achieve the effect of real-time temperature control of the grinder system.

[0034] It should be noted that the above scheme is mainly for the thermal management control of a single grinder. For the combined control of multiple grinders, please refer to Figure 3 at the same time. Set 2s devices, divided into two working conditions: rough grinding s and fine grinding s. During the trial production process, adjust the two working conditions respectively, and collect the adjusted T and ω, where is the start-up time of the device, is the production cycle, the value of and the value of represent the rough grinding working condition, the value of and the value of represent the fine grinding working condition, represents the cooling water input of the i-th device during rough grinding, represents the cooling water input of the i-th device during fine grinding, T, ω, 、 all follow the Gaussian distribution 、 、 、 , and .

[0035] The PSO algorithm is used for optimization. The PSO algorithm is the particle swarm optimization algorithm. Set the total water consumption as G, and the cooling water input of each device as Q. For the rough grinding condition value , and for the fine grinding condition value . The PSO algorithm discretizes the function G, that is, discretely samples the rough grinding stage and the fine grinding stage, reducing the system's requirement for computing power. Since the fluctuation of the input cooling water during the grinding operation of the equipment is large, the sampling discretization is based on the Gaussian distribution function, ensuring that the sampling density is high in the early stage of equipment startup and decreases after stabilization in the later stage. The sampling formula of the Gaussian distribution function is as follows: ; where m represents the sampling point. For example, set M sampling points, ; is the corresponding time point; are all fitting parameters obtained during the actual project debugging, w 1 represents the peak height of the Gaussian function, that is, the maximum value of the function at the center position, w 2 corresponds to the mean of the Gaussian distribution, that is, the position of the distribution symmetry axis, w 3 is related to the variance, can be taken as 51660, 51.08, and 17.1 respectively.

[0036] After determining the sampling point positions, the linear interpolation method is used to select the and corresponding to M sampling points, and combined with the square wave convolution function , a function G is formed after convolution with the square wave function. The square wave function sets the time difference based on the different water demand in different time periods: ; The PSO algorithm is used to optimize the function G to obtain the optimized T value. At the same time, record the actual production situation, record the actual production cycle and start / stop time of each device, return the earliest updated T value and ω value, and avoid system overfitting. After several iterations, a T value that is more suitable for the actual joint control of the engineering project is formed.

[0037] In the case of cross-operating conditions, it is necessary to consider the T values and ω values of the same equipment at different times and under different operating conditions simultaneously. Based on the component conditions and actual working states of each equipment, there will be differences. Since it is cross-operating, when statistically analyzing the Gaussian distribution, the two production lines are not statistically analyzed separately. Instead, data fusion is performed on the T values and ω values of each equipment to obtain 、 . Meanwhile, is the cooling water input of the i-th equipment during rough grinding, is the cooling water input of the i-th equipment during fine grinding. Similarly, data fusion is performed on the two production lines, which follows the Gaussian distribution .

[0038] Similarly, using the PSO algorithm combined with the Gaussian process, the total water consumption is set as G: ; The PSO algorithm is used to optimize the above G function to obtain the optimized T values and ω values. Similarly, after several iterations, a T value that is more suitable for the actual joint control of the engineering project is formed.

[0039] Therefore, for the joint control of multiple grinders, especially during cross-operating condition adjustment and optimization, this method uses data fusion to perform data fusion on the production processes of multiple production lines, reducing the computing cost. Through optimal control, the water inlet cycle under different operating conditions is adjusted, avoiding the problem of multiple equipment competing for cooling water simultaneously, resulting in too high a temperature rise, and the problem of insufficient water consumption of the cooling water under certain operating conditions.

[0040] The following further illustrates this method through a specific case. By comparing this method with PI control respectively, the rod pin type nano grinder PHN SuperMaxZeta 400 is selected as the grinder for this test. The project is carried out at the workshop site of a new energy cathode material production factory in a certain place. The same 0.3mm zirconia beads are used as the grinding medium, with a filling rate of about 80% and a grinding time of 90 minutes. With the same rotor structure, current and rotational speed (linear speed), the same flow rate and temperature, the calculation results using simulink in matlab show that both can control the outlet temperature of the grinding slurry at about 12°C and the outlet temperature of the cooling water below 30°C in a relatively short time. However, it can be seen that this method reaches stability faster than PID control. In addition, the cooling water consumption using this method can also reach balance faster.

[0041] For 2 sFor the joint control of multiple devices with [[ID=]] = 12, 28 sampling points are set. When manually adjusting the start and stop of multiple devices, there will be a situation where the devices compete for cooling water at the initial stage of operation. The designed cooling water margin needs to be increased. Through the coupling adjustment of this method (PSO-GP, particle swarm optimization algorithm combined with Gaussian process), it has a significant improvement compared to manual and PSO algorithms, has less impact on the total amount of cooling water, and avoids the problem of uneven distribution of cooling water at the initial stage of operation.

[0042] The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity of description, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered as the scope described in this specification.

[0043] The above embodiments only represent several implementation manners of the present application, and their descriptions are relatively specific and detailed, but they should not be construed as limiting the scope of the application. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present application, several modifications and improvements can still be made, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the appended claims.

Claims

1. A finite element based MPC grinding machine thermal management control method, characterized in that: The steps include: S1, establish a physical model related to the thermal management of the grinding machine system and use the finite element simulation data as the prediction sample space; S2, according to the delay of the grinding machine system temperature, the target transfer function of the prediction model control is established as the control equation of the model prediction; S3, rolling optimization and feedback correction of the prediction model are performed through the MPC control algorithm, and the predicted temperature change trend is integrated with the real data measured by the temperature sensor to obtain the optimal output change after integration; S4, determine the input flow rate according to the optimal output change to achieve the effect of real-time temperature control of the grinding machine system; for the joint control of multiple grinding machines, set 2 s Equipment, divided into coarse grinding s Table and fine grinding s During the trial production, the two working conditions were adjusted and the adjusted T and ω ,in is the startup time of the device, For the production cycle, Value and Value Represents the rough grinding condition, Value and Value Represents fine grinding conditions, represents the cooling water input of the i-th equipment during rough grinding, represents the cooling water input of the i-th equipment during fine grinding, T, ω, , All obey Gaussian distribution , , , , and ; The PSO algorithm is used for optimization. The total water consumption is set to G, the cooling water input of each device is set to Q, and the rough grinding condition is set to Value , fine grinding conditions Value , discretize the function G, and the sampling discretization is based on the Gaussian distribution function, so that the sampling density is high in the early stage of the device startup and the sampling density is reduced after the device stabilizes in the later stage; after determining the sampling point position, the linear interpolation method is used to select the M sampling points corresponding to and , and then convolved with the square wave function to form function G. The square wave function sets the time difference based on the different water demand in different time periods. The PSO algorithm is optimized to obtain the optimized T value. At the same time, the actual production situation is recorded. The actual production cycle and start and stop time of each equipment are recorded, and the earliest updated T value and ω value are returned. After several iterations, the T value of joint control is obtained.

2. The MPC grinding machine thermal management control method based on finite element according to claim 1 is characterized in that: The physical model established in S1 is: ； in, is the specific heat capacity of the grinding slurry, in units of ; is the specific heat capacity of cooling water, in units of ; is the outlet temperature of the grinding slurry, in units of ; is the outlet temperature of cooling water, in ; is the density of the grinding slurry, in units of ; is the density of cooling water, in units of ; is the volume inside the grinder, in ; is the volume of the cooling jacket, in ; is the input flow rate of the grinding slurry, in units of ; is the input flow rate of cooling water, in units of ; is the inlet temperature of the grinding slurry, in units of ; is the cooling water inlet temperature, in ; W is the heat generated by the grinder, in units of ; U is the heat transfer coefficient, unit is ; S is the heat transfer area, unit is .

3. The MPC grinding machine thermal management control method based on finite element according to claim 2 is characterized in that: The heat generation W of the grinding machine in S1 is taken as the input value by the result of finite element simulation, and the input flow rate of the grinding slurry is Set to a fixed value to adjust the cooling water input flow Set as input value to set the outlet temperature of the grinding slurry and cooling water outlet temperature Set to output value.

4. The MPC grinding machine thermal management control method based on finite element according to claim 3 is characterized in that: The target transfer function in S2 for: ； in, is the output matrix of the system, including the outlet temperature of the polishing slurry and the cooling water outlet temperature ; is the input value of the system, that is, the input flow rate of cooling water ; is the output weight matrix of the system, , is the weight factor of cooling water temperature, is the weight factor of the grinding slurry temperature; is the weight factor of cooling water flow.

5. The MPC grinding machine thermal management control method based on finite element according to claim 4 is characterized in that: According to the needs of the actual project, use different weight factors to adjust the weights between cooling water and grinding slurry. The value of increases, When the value decreases, the focus is on controlling the temperature. The value of When the value increases, the temperature is controlled under the premise of saving flow.

6. The MPC grinding machine thermal management control method based on finite element according to claim 5 is characterized in that: S3 specifically includes the following sub-steps: S31, in each sampling cycle, the inlet temperature and outlet temperature of the grinding machine system at adjacent moments are optimized to find the optimal control sequence of the predicted value, so as to predict the outlet temperature value of the grinding machine system at the next moment, and the controlled variable of the controlled object is obtained according to the outlet temperature predicted value at the next moment, and the next prediction initialization is performed, and the iteration is continued with the optimization time; S32, during each sampling cycle, the temperature trend error generated by the prediction model is compensated by collecting the outlet temperature measurement value of the grinder in real time.

7. The MPC grinding machine thermal management control method based on finite element according to claim 1 is characterized in that: The sampling formula of Gaussian distribution function is: ； in, m Indicates the sampling point, set M sampling points, ; is the corresponding time point; These are all fitting parameters, obtained during actual project debugging. w 1 represents the peak height of the Gaussian function, that is, the maximum value of the function at the center. w 2 corresponds to the mean of the Gaussian distribution, that is, the position of the symmetry axis of the distribution, w 3 is related to the variance.

8. The MPC grinding machine thermal management control method based on finite element according to claim 1 is characterized in that: function Specifically: 。 9. A finite element based MPC grinding machine thermal management control method according to any one of claims 1 to 6, characterized in that: For the cross-operating joint control of multiple grinding machines, the T value and ω value of each device are fused when the Gaussian distribution is statistically analyzed. is the startup time of the device, is the production cycle, which follows Gaussian distribution , , is the cooling water input of the i-th equipment during rough grinding, is the cooling water input of the i-th equipment during fine grinding, which follows a Gaussian distribution after fusion ; The sampling formula of Gaussian distribution function is: ； Among them, m represents the sampling point, and M sampling points are set. ; is the corresponding time point; These are all fitting parameters, obtained during actual project debugging. w 1 represents the peak height of the Gaussian function, that is, the maximum value of the function at the center. w 2 corresponds to the mean of the Gaussian distribution, that is, the position of the symmetry axis of the distribution, w 3 is related to variance; Using the PSO algorithm combined with the Gaussian process, the total water consumption is set to : ； The G function is optimized by PSO algorithm to obtain the optimized T value and ω value. After several iterations, the T value of joint control is obtained.

Images