Control method and system based on refrigerator shell production line

Abstract

The present application relates to the control adjustment technical field, especially control method and system based on refrigerator shell production line, method includes: obtaining the power value and current value of injection molding motor for controlling refrigerator shell production line at each time; obtain the adaptive loudness attenuation coefficient before each output of bat algorithm optimal solution for controlling the power value of injection molding motor, including: determining the power data reference intensity and power data true intensity of the time period to which each time belongs, to obtain the adaptive loudness attenuation coefficient; according to the energy consumption and product pass rate of each candidate solution before each output of optimal solution, construct objective function; based on adaptive loudness attenuation coefficient and objective function, filter each output of optimal solution, realize the control of refrigerator shell production line. The present application collects injection molding motor data in real time, dynamically adjusts the loudness attenuation coefficient of bat algorithm, thereby improving the stability, efficiency and product quality of refrigerator shell production line.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of control adjustment technology, in particular to a control method and system based on a refrigerator shell production line. BACKGROUND

[0002] With the continuous evolution of industrial automation and intelligent manufacturing, advanced intelligent control technology is gradually introduced into traditional manufacturing fields to improve production efficiency, reduce energy consumption and optimize product quality. In the household appliance manufacturing industry, as a common and key consumer product, the efficiency and quality of the production process of refrigerators directly affect the market competitiveness and resource utilization level of enterprises. Especially in the refrigerator shell production line, the injection molding process occupies a core position, and the working state of the injection molding motor directly affects the production efficiency, energy consumption and the quality of the final product: if the power of the injection molding motor is insufficient, it may cause the injection molding cycle to be prolonged and the production speed to be slowed down; if the power of the injection molding motor is too high, it will cause energy waste and may also cause the risk of equipment overload, thereby affecting the service life of the equipment and the stability of the production line.

[0003] As an intelligent optimization algorithm simulating the echolocation behavior of bats in nature, the bat algorithm has strong global search capability and can effectively handle complex optimization problems, so that the injection molding motor can maintain stable output under different process conditions, avoid production delays and quality problems caused by power fluctuations, and improve production efficiency and product pass rate.

[0004] However, when the bat algorithm outputs the optimal solution of the power data of the injection molding motor, a fixed value of the loudness attenuation coefficient is set. The loudness attenuation coefficient and the initial loudness jointly determine the loudness value corresponding to each iteration in the optimal solution output process, and the loudness value and the quality of the new candidate solution obtained in each iteration process jointly affect whether the new candidate solution is accepted. Therefore, when the fixed loudness attenuation coefficient is set too large, the bat algorithm may continuously produce large disturbances and it is difficult to achieve convergence; when the fixed loudness attenuation coefficient is set too small, the bat algorithm may lock the local optimal solution too early and cannot complete the search for the global optimal solution, thereby affecting the production stability, energy utilization efficiency and final product quality of the refrigerator shell production line. SUMMARY

[0005] In order to solve the problem that the bat algorithm sets a fixed value of the loudness attenuation coefficient, when the fixed loudness attenuation coefficient is set too large, the bat algorithm may continuously produce large disturbances and it is difficult to achieve convergence, and when the fixed loudness attenuation coefficient is set too small, the bat algorithm may lock the local optimal solution too early and cannot complete the search for the global optimal solution, thereby affecting the production stability, energy utilization efficiency and final product quality of the refrigerator shell production line, the present application provides a control method and system based on a refrigerator shell production line.

[0006] In a first aspect, the present application provides a refrigerator shell production line control method, which adopts the following technical solution:

[0007] The refrigerator shell production line control method comprises: based on the collected power value and current value of the injection molding motor for controlling the refrigerator shell production line at each time, obtaining the adaptive loudness decay coefficient before each output optimal solution of the bat algorithm for controlling the power value of the injection molding motor, constructing a target function according to the energy consumption and product pass rate of each candidate solution before each output optimal solution, and screening the optimal solution output each time based on the adaptive loudness decay coefficient and the target function to realize the control of the refrigerator shell production line. In the process of obtaining the adaptive loudness decay coefficient, the optimal solution output at any time is recorded as a target sub-optimal solution, the time corresponding to the output of the target sub-optimal solution is recorded as a target time, the power data reference intensity of the time period to which the target time belongs is determined according to the difference between the power values of all adjacent two times in the time period and the information entropy of the power values of all times in the time period to which the target time belongs; the power data real intensity of the time period to which the target time belongs is determined according to the power value of each time in the time period in the power value of all times in the time period to which the target time belongs and the power value of each time in the time period in the power value of all times in the time period to which the target time belongs; and the adaptive loudness decay coefficient before the target sub-optimal solution is obtained by modifying the preset loudness decay coefficient of the bat algorithm according to the power data real intensity.

[0008] The beneficial effects are that by collecting the power value and current value of the injection molding motor in real time, combining the reference intensity and real intensity of the power data in the time period, dynamically adjusting the loudness decay coefficient of the bat algorithm, the convergence problem of the algorithm caused by improper setting of the traditional fixed coefficient is overcome; the adaptive loudness decay coefficient can automatically adjust according to the actual working condition of the production process, appropriately reduce the decay coefficient to enhance the local search ability when the power data intensity is high to avoid large disturbance; appropriately increase the decay coefficient to maintain the global search ability when the power data intensity is low to prevent falling into a local optimal solution; by constructing the target function containing the energy consumption and product pass rate, the energy efficiency and product quality requirements of the production process are comprehensively considered, so that the selection of the optimal solution is more scientific and reasonable; the synergistic optimization based on the adaptive loudness decay coefficient and the target function can effectively balance the convergence speed and search accuracy of the algorithm, improve the control stability of the refrigerator shell production line; avoids the subjectivity and limitation of manually setting fixed parameters, realizes the intelligentization and adaptability of parameter adjustment, significantly improves the production stability, energy utilization efficiency and product quality consistency of the refrigerator shell production line, and provides effective technical support for the intelligent control of industrial production processes.

[0009] Further, the time period to which the target moment belongs is a time period constituted by the target moment and a plurality of moments before the target moment.

[0010] Further, the power data reference intensity of change satisfies: a sum of differences between power values of all adjacent moments in a time period to which a target moment belongs and a product of a normalized result of a multiplication of an information entropy of power values of all moments in the time period to which the target moment belongs, is recorded as a power data reference intensity of change of the time period to which the target moment belongs.

[0011] The beneficial effect is that: by calculating the product of the sum of differences between power values of adjacent moments in the time period to which the target moment belongs and the information entropy of power values, and performing normalization processing, a quantitative index of the power data reference intensity of change is constructed, the sum of differences embodies the overall amplitude of power change, the information entropy reflects the discrete degree of power distribution, the product of the two makes the intensity of change evaluation more comprehensive and accurate, the normalization processing eliminates the influence of different dimensions and orders of magnitude, facilitating comparison and analysis between data of different time periods, providing a reliable basis for subsequent adaptive loudness decay coefficient adjustment, the power data reference intensity of change comprehensively considers the amplitude feature and distribution feature of power change, and can effectively reflect the fluctuation characteristics of power data.

[0012] Further, the power data true intensity of change satisfies: a sum of differences between a percentile of a power value of each moment in a time period to which a target moment belongs in power values of all moments in the time period and a percentile of a current value of each moment in the time period to which the target moment belongs in power values of all moments in the time period is recorded as a correction factor of the target moment; a normalized result of a negative correlation mapping of the power data reference intensity of change and the correction factor is adjusted based on a first adjustment coefficient and a second adjustment coefficient used for adjusting a value range of a result, to obtain a power data true intensity of change of the time period to which the target moment belongs.

[0013] The beneficial effect is that: by calculating the sum of differences between the percentiles of the power value and the current value of each moment in the time period to which the target moment belongs, a correction factor is constructed to reflect the relative change characteristics of power data, the correction factor can effectively capture the coordinated change between power and current, avoiding the one-sidedness of single power index evaluation; by the negative correlation mapping of the power data reference intensity of change and the correction factor, and the range adjustment of the first adjustment coefficient and the second adjustment coefficient, the accurate quantification of the power data true intensity of change is realized; the negative correlation mapping mechanism makes the true intensity of change decrease when the correction factor is large (power and current change are not coordinated), and vice versa, and this adjustment mechanism is more in line with the physical law of the actual production process; the design of the double adjustment coefficients ensures that the value range of the true intensity of change is reasonable, avoids the influence of extreme values on the algorithm performance, and improves the accuracy and stability of the adaptive loudness decay coefficient adjustment.

[0014] Further, the self-adaptive loudness attenuation coefficient is a product of the real intensity of the power data and a preset loudness attenuation coefficient of the bat algorithm, and is recorded as a self-adaptive loudness attenuation coefficient before a target sub-optimal solution.

[0015] The beneficial effect is that by multiplying the real intensity of the power data and the preset loudness attenuation coefficient of the bat algorithm, a simple and efficient self-adaptive loudness attenuation coefficient calculation model is constructed, so that the real intensity can directly adjust the size of the preset coefficient, realizing the intuitiveness and interpretability of parameter adjustment; when the real intensity of the power data is high, the self-adaptive loudness attenuation coefficient is correspondingly increased, enhancing the global search ability of the algorithm and avoiding premature convergence, and when the real intensity of the power data is low, the self-adaptive loudness attenuation coefficient is correspondingly reduced, improving the local search precision of the algorithm and accelerating the convergence speed.

[0016] Further, the target function satisfies that a sum of a normalized result of the energy consumption of each candidate solution before the optimal solution of each output and a negative correlation mapping of the product pass rate of each candidate solution before the optimal solution of each output is recorded as a result of the target function of each candidate solution before the optimal solution of each output.

[0017] The beneficial effect is that by adding the normalized result of the energy consumption of the candidate solution and the negative correlation mapping of the product pass rate, a target function considering energy efficiency and product quality is constructed; the energy consumption normalization eliminates the dimensional difference under different production conditions, so that the energy consumption index is comparable; the negative correlation mapping of the product pass rate reflects the optimization direction that the higher the pass rate, the smaller the target function value, which meets the optimization goal of minimization; by considering two key indicators at the same time, the problem of losing one side to gain the other side caused by single indicator optimization is avoided, so that the optimal solution selected realizes reasonable control of energy consumption under the premise of ensuring product quality, and improves the overall economic benefit and production efficiency of the refrigerator shell production line.

[0018] Further, the energy consumption is the energy consumption within a preset time when the candidate solution is applied to control the refrigerator shell production line.

[0019] Further, the product pass rate is the product pass rate in the historical refrigerator shell detection data obtained in advance when the candidate solution is applied to control the refrigerator shell production line.

[0020] Further, the screening the optimal solution comprises: based on the initial loudness of the adaptive loudness decay coefficient and the bat algorithm, obtaining a loudness value corresponding to each round of iteration when the bat algorithm outputs an optimal solution each time; in each round of iteration, in response to the random number generated by the bat algorithm being less than the loudness value and the target function result of the new candidate solution obtained in the current round of iteration being less than the candidate solution when the last round of iteration is completed, the new candidate solution obtained in the current round of iteration is taken as the candidate solution when the current round of iteration is completed, otherwise, the candidate solution when the current round of iteration is completed is the candidate solution when the last round of iteration is completed; and the candidate solution when the last round of iteration is completed is taken as the optimal solution output by the bat algorithm each time.

[0021] In a second aspect, the present application provides a control system based on a refrigerator shell production line, which adopts the following technical scheme:

[0022] The control system based on the refrigerator shell production line comprises a processor and a memory, and the memory stores computer program instructions.

[0023] By adopting the above technical scheme, the control method based on the refrigerator shell production line is generated into a computer program and stored in the memory to be loaded and executed by the processor, so that a terminal device is manufactured according to the memory and the processor, and use is facilitated.

[0024] The present application has the following technical effects:

[0025] (1) Breakthrough the limitation of the traditional bat algorithm fixed loudness decay coefficient, dynamically correct the decay coefficient through the power data intensity, realize adaptive adjustment, when the power data fluctuates violently (such as production line load mutation), the adaptive coefficient can inhibit excessive disturbance, avoid algorithm divergence; when the data is stable, the coefficient can promote the deepening of local search, prevent premature falling into local optimum, significantly improve the algorithm convergence stability and global optimal solution search ability, ensure that the output injection motor power optimal solution is more in line with the actual production demand.

[0026] (2) By fusing the power value difference, information entropy and the percentage of current and power, etc. Multi-dimensional features, accurately describe the real intensity of power data, make the adjustment of loudness decay coefficient highly match the actual working condition of production line (such as raw material batch difference, equipment wear degree, environmental temperature change, etc.); for example, in the melting stage and cooling stage of refrigerator shell injection molding, the power fluctuation characteristics are different, the adaptive coefficient can optimize the search strategy, reduce the motor power fluctuation caused by parameter mismatch, so as to ensure the stability of production line operation and reduce the equipment failure rate.

[0027] (3) The adaptive mechanism makes the algorithm not need to rely on artificial preset fixed parameters, can independently adapt to the data distribution change in the long-term operation of the production line (such as the power baseline drift caused by equipment aging), not only reduces the workload and error of artificial debugging, but also improves the fault tolerance of the control system to complex working conditions, ensures the continuous output of stable and reliable control strategies in the dynamic production environment, and provides technical support for the intelligent upgrading of the refrigerator shell production line. BRIEF DESCRIPTION OF DRAWINGS

[0028] Figure 1 is a method flowchart in the control method of the refrigerator shell production line based on the embodiment of the application.

[0029] Figure 2 is a method flowchart of step S2 in the control method of the refrigerator shell production line based on the embodiment of the application. DETAILED DESCRIPTION

[0030] The technical solutions in the embodiments of the application will be clearly and completely described below with reference to the drawings in the embodiments of the application. Obviously, the described embodiments are part of the embodiments of the application, rather than all the embodiments of the application. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the application.

[0031] The embodiment of the application discloses a control method based on a refrigerator shell production line, referring to Figure 1 , comprising steps S1-S6:

[0032] S1: Obtain the power value and current value of the injection molding motor for controlling the refrigerator shell production line at each time.

[0033] It should be noted that the application will use a power acquisition instrument and a current sensor to collect power data and current data of the injection molding motor. The collection of the two kinds of data needs to be synchronized, and the collection frequency of the above two kinds of data is 10Hz. Then, an analog-to-digital conversion device is used to perform analog-to-digital conversion on the above two kinds of data to obtain the digital representation of the above two kinds of data for subsequent step analysis.

[0034] S2: Obtain the adaptive loudness attenuation coefficient before the optimal solution of the power value for controlling the injection molding motor is output by the bat algorithm each time.

[0035] It should be noted that the core purpose of the application is to use the bat algorithm to intelligently control the power data of the injection molding motor. The loudness attenuation coefficient of the bat algorithm each time the optimal solution of the power data of the injection molding motor is output is adapted, which is mainly based on the analysis of the numerical change characteristics of the power data and the current data of the injection molding motor.

[0036] Based on the collected power value and current value of the injection motor for controlling the refrigerator shell production line at each moment, the adaptive loudness decay coefficient before each output of the bat algorithm optimal solution for controlling the power value of the injection motor is obtained.

[0037] Specifically, the time period of the target moment is a time period composed of the target moment and a plurality of moments before the target moment.

[0038] The number of the plurality of moments before the target moment can be set by the implementer according to the specific implementation, for example, 300, and for the first 30 seconds of data collected, only the time period of the moment is used to construct, and the following calculation is not performed.

[0039] Reference Figure 2 In the process of obtaining the adaptive loudness decay coefficient, steps S201-S203 are included, and the details are as follows:

[0040] S201: Determine the power data reference intensity of the time period of each moment.

[0041] It should be noted that the core logic of the present scheme is to adaptively adjust the loudness decay coefficient for each optimal solution output of the power data of the injection motor according to the true intensity of the power data of the injection motor in the time period of the corresponding moment when the optimal solution of the power data of the injection motor is output each time. Then, the value change characteristics of the power data of the injection motor are obtained in this step to obtain the power data reference intensity of the time period of the corresponding moment of the injection motor. When analyzing this index, if the power data of the injection motor fluctuates more frequently and the numerical difference is more obvious when the optimal solution of the power data of the injection motor is output each time, it can be concluded that the power data of the injection motor fluctuates more intensely in the time period of the moment, and the power data reference intensity of the time period of the corresponding moment corresponding to each optimal solution output of the power data of the injection motor is greater.

[0042] The optimal solution output at any time is recorded as the target sub-optimal solution, the moment corresponding to the output of the target sub-optimal solution is recorded as the target moment, and the power data reference intensity of the time period of the target moment is determined according to the difference between the power values of all adjacent two moments in the time period of the target moment and the information entropy of the power values of all moments in the time period of the target moment.

[0043] Specifically, the power data reference intensity satisfies:

[0044] , The power data reference intensity of the time period of the moment corresponding to the output of the first optimal solution is ​For the first The optimal solution output is the number of times within the time period corresponding to the time of output. and The first The optimal solution for the output is the [number]th [time]th [time] within the time period corresponding to the time of output. The moment and the first The power value at each moment. For the first The optimal solution output is the information entropy of the power values ​​at all times within the time period corresponding to the time of output. For the standard normalized function, It is the absolute value symbol.

[0045] in, The larger it is, the more likely it is to be the first The more significant the difference in power values ​​between all adjacent times within the time period corresponding to the optimal solution of the output, the more it indicates that the optimal solution of the second output is... The optimal solution output is most relevant to the power data intensity of the time period at which it occurs. However, due to the... The optimal solution for this output may exhibit regular (similar amplitude) fluctuations in power values ​​within the time period corresponding to the output time. Therefore, it is calculated in conjunction with information entropy. The larger it is, the more it indicates the number of... The more chaotic the power value distribution within the time period corresponding to the optimal solution of the second output, the more it can be explained that the second optimal solution is the most chaotic. The more drastic the changes in data at each time point within the time period corresponding to the optimal solution output, the greater the reliability, and the greater the drastic the reference power data.

[0046] S202: Determine the true intensity of power data for each time period.

[0047] It should be noted that the power data reference intensity is based on the numerical change characteristic analysis of the power data of the injection molding motor, but when collecting the power data of the injection molding motor, other electrical equipment (such as high-power electrical equipment, compressors, etc.) may generate electromagnetic wave interference, thereby affecting the accuracy of the power data, leading to the generation of power noise data, and the power noise data may be close to the power real data in the numerical level. Therefore, it is necessary to denoise the power data reference intensity to obtain the real distance degree. Then, according to the scene investigation, there is a strong correlation between the power data and the current data of the injection molding motor, and the current of the motor is directly proportional to its load. When the load of the injection molding machine increases (for example, the mold pressure increases during the injection molding process), the current will increase. Since the power is proportional to the current, the power will also increase with the increase of the current. Therefore, in this step, the numerical change characteristics of the power data and the current data of the injection molding motor are analyzed, and the power data real intensity of the injection molding motor is obtained by optimizing the power data reference intensity of the injection molding motor. The logic is as follows: the stronger the correlation between the power value and the current value in the time period to which the moment corresponding to each time the optimal solution of the power data of the injection molding motor is output belongs, the less likely it is that the power data of the time period to which the moment corresponding to each time the optimal solution of the power data of the injection molding motor is output belongs is affected by noise, and then it can be indicated that the greater the power data reference intensity, the greater the corresponding power data real intensity.

[0048] According to the power data reference intensity, the percentile of the power value of each time in the time period to which the target moment belongs among all power values in the time period, and the percentile of the current value of each time in the time period to which the target moment belongs among all power values in the time period, the power data real intensity of the time period to which the target moment belongs is determined.

[0049] Specifically, the power data real intensity satisfies:

[0050] , and are the power data real intensity and the power data reference intensity of the time period to which the moment corresponding to the output of the optimal solution of the i-th time belongs, respectively, is the number of times in the time period to which the moment corresponding to the output of the optimal solution of the i-th time belongs, is the power value of the j-th time in the time period to which the moment corresponding to the output of the optimal solution of the i-th time belongs, is the power value of the j-th time in the time period to which the moment corresponding to the output of the optimal solution of the i-th time belongs, is the percentile of the power value of the j-th time in the time period to which the moment corresponding to the output of the optimal solution of the i-th time belongs among all power values in the time period, is the percentile of the power value of the j-th time in the time period to which the moment corresponding to the output of the optimal solution of the i-th time belongs among all power values in the time period, is the percentile of the power value of the j-th time in the time period to which the moment corresponding to the output of the optimal solution of the i-th time belongs among all power values in the time period, is the percentile of the power value of the j-th time in the time period to which the moment corresponding to the output of the optimal solution of the i-th time belongs among all power values in the time period, the power data real drastic degree of the time period to which the time corresponding to the optimal solution output by the i-th output corresponds, the percentile of the current value of the time corresponding to the optimal solution output by the i-th output in the current values of all times in the time period, and are respectively a first adjustment coefficient and a second adjustment coefficient for adjusting the value range of the result, is a natural exponential function, is an absolute value symbol.

[0051] The first adjustment coefficient and the second adjustment coefficient can be set by the implementer according to the specific situation, for example, the first adjustment coefficient is 0.2 and the second adjustment coefficient is 0.9, and finally the value range of the calculation result is adjusted to the interval by the first adjustment coefficient and the second adjustment coefficient.

[0052] wherein, The greater the power data real drastic degree of the time period to which the time corresponding to the optimal solution output by the i-th output corresponds, the greater the power data real drastic degree of the time corresponding to the optimal solution output by the i-th output. indicates the correlation between the power data and the current data in the time period to which the time corresponding to the optimal solution output by the i-th output corresponds, the smaller the value, the stronger the correlation between the power data and the current data in the time period to which the time corresponding to the optimal solution output by the i-th output corresponds, the smaller the possibility that the power data in the time period to which the time corresponding to the optimal solution output by the i-th output corresponds is affected by noise, and the greater the credibility of

[0053] S203: Determine the adaptive loudness decay coefficient before the optimal solution output by the bat algorithm each time for controlling the power value of the injection molding motor.

[0054] It should be noted that after obtaining the power data real drastic degree, the loudness decay coefficient before the optimal solution output by the bat algorithm each time for controlling the power value of the injection molding motor is adaptively adjusted. The logic is that the greater the power data real drastic degree, the more unstable the injection molding motor is, in order to avoid premature convergence of the algorithm and miss the optimal solution, the bat algorithm needs to set a larger loudness decay coefficient to slow down the decay rate of the loudness and retain more random disturbance; on the contrary, the smaller the power data real drastic degree, the more stable the injection molding motor is, in order to reduce invalid disturbance, the bat algorithm needs to set a smaller loudness decay coefficient to accelerate the decay of the loudness and quickly lock the optimal solution.

[0055] ​​​​​Based on the actual intensity of the power data, the loudness attenuation coefficient preset by the Bat Algorithm is corrected to obtain the adaptive loudness attenuation coefficient before the target suboptimal solution.

[0056] Specifically, the adaptive loudness attenuation coefficient:

[0057] , For the first The adaptive loudness attenuation coefficient before the optimal solution of the next output. The loudness attenuation coefficient preset for the bat algorithm, For the first The optimal solution output is the power data of the time period corresponding to the time of output, which reflects the actual intensity of the power surge.

[0058] The implementers can set the loudness attenuation coefficient according to the specific implementation situation, for example, 0.9.

[0059] S3: Construct an objective function based on the energy consumption and product qualification rate of each candidate solution before the optimal solution output each time.

[0060] Specifically, the energy consumption refers to the energy consumption within a preset time when the candidate solution is applied to control the refrigerator shell production line.

[0061] The implementers can set a preset time according to the specific implementation situation, for example, 10 seconds.

[0062] Specifically, the product qualification rate is the product qualification rate in the pre-acquired historical refrigerator shell inspection data when applying candidate solutions to control the refrigerator shell production line.

[0063] Specifically, the objective function satisfies:

[0064] , For the first The first optimal solution output before the [number]th ... The result of the objective function of each candidate solution. For the first The first optimal solution output before the [number]th ... The energy consumption of each candidate solution. For the first The first optimal solution output before the [number]th ... Product qualification rate of each candidate solution. For the standard normalized function, It is a natural exponential function.

[0065] S4: Based on the adaptive loudness attenuation coefficient and objective function, the optimal solution for each output is selected to realize the control of the refrigerator shell production line.

[0066] It should be noted that the algorithm will randomly generate several bats (initial solutions) at the beginning, calculate the objective function of each bat (in the same way as the calculation of the objective function of the candidate solution, which will not be described here), obtain the initial candidate solution, and then perform the subsequent iteration process.

[0067] Specifically, the optimal solution is screened, including:

[0068] Based on the adaptive loudness decay coefficient and the initial loudness of the bat algorithm, the loudness value corresponding to each iteration of the bat algorithm is obtained when the optimal solution is output each time.

[0069] In each iteration process, in response to the random number generated by the bat algorithm being less than the loudness value, and the objective function result of the new candidate solution obtained in the current iteration being less than the candidate solution at the end of the last iteration, the new candidate solution obtained in the current iteration is taken as the candidate solution at the end of the current iteration, otherwise, the candidate solution at the end of the current iteration is the candidate solution at the end of the last iteration.

[0070] The candidate solution at the end of the last iteration when the iteration process ends is taken as the optimal solution output by the bat algorithm each time.

[0071] The implementer can adjust the parameters of the bat algorithm according to the specific implementation, for example, the population size (number of bats) is preset to an empirical value of 30, the solution space is preset to an empirical value of , the speed of the solution space range is preset to an empirical value of , the maximum number of iterations is preset to an empirical value of 60, the frequency range is preset to an empirical value of , the initial loudness is preset to an empirical value of 1, the initial pulse rate is preset to an empirical value of 0.2, and the pulse enhancement coefficient is preset to an empirical value of 0.9.

[0072] The embodiment of the application also discloses a control system based on a refrigerator shell production line, comprising a processor and a memory, and the memory stores computer program instructions, which realize the control method based on the refrigerator shell production line according to the application when executed by the processor.

[0073] The above system also includes a communication bus and a communication interface and other components familiar to those skilled in the art, the settings and functions of which are known in the art, and therefore will not be described here.

[0074] The above are the preferred embodiments of the application, which do not limit the protection scope of the application, therefore: any equivalent changes made on the basis of the structure, shape, principle of the application should be covered within the protection scope of the application.

Claims

1. Control method based on a refrigerator shell production line, characterized in that, The method comprises the following steps: Based on the power value and the current value of the injection molding motor for controlling the refrigerator shell production line at each moment, an adaptive loudness decay coefficient before each output optimal solution of the bat algorithm for controlling the power value of the injection molding motor is obtained, a target function is constructed based on the energy consumption and the product qualification rate of each candidate solution before each output optimal solution, and the optimal solution output each time is screened based on the adaptive loudness decay coefficient and the target function, so as to realize the control of the refrigerator shell production line. In the process of obtaining the adaptive loudness decay coefficient, the optimal solution output at any time is recorded as a target sub-optimal solution, the moment corresponding to the output of the target sub-optimal solution is recorded as a target moment, the power data reference intensity of the time period to which the target moment belongs is determined according to the difference between the power values of all adjacent moments in the time period to which the target moment belongs and the information entropy of the power values of all moments in the time period to which the target moment belongs; the power data real intensity of the time period to which the target moment belongs is determined according to the power value of each moment in the time period to which the target moment belongs, the percentile of the power value of each moment in the time period to which the target moment belongs in all moments in the time period, and the percentile of the current value of each moment in the time period to which the target moment belongs in all power values in the time period; and the adaptive loudness decay coefficient before the target sub-optimal solution is obtained by correcting the preset loudness number decay coefficient of the bat algorithm according to the power data real intensity.

2. The control method of the refrigerator case production line according to claim 1, characterized by, The time period to which the target moment belongs is a time period composed of the target moment and a plurality of moments before the target moment.

3. The control method of the refrigerator case production line according to claim 1, characterized by, The power data reference intensity satisfies: The power data reference intensity of the time period to which the target moment belongs is the normalized result of the product of the sum of the differences between the power values of all adjacent moments in the time period to which the target moment belongs and the information entropy of the power values of all moments in the time period to which the target moment belongs.

4. The control method of a refrigerator case production line according to claim 1, characterized by, The power data real intensity satisfies: The difference between the percentile of the power value of each moment in the time period to which the target moment belongs in all power values in the time period and the percentile of the current value of each moment in the time period to which the target moment belongs in all power values in the time period is recorded as a correction factor of the target moment. The power data real intensity of the time period to which the target moment belongs is obtained by adjusting the normalized result of the negative correlation mapping of the power data reference intensity and the correction factor based on a first adjustment coefficient and a second adjustment coefficient for adjusting the value range of the result.

5. The control method of a refrigerator case production line according to claim 1, characterized by, The adaptive loudness decay coefficient satisfies: The product of the power data real intensity and the preset loudness number decay coefficient of the bat algorithm is recorded as the adaptive loudness decay coefficient before the target sub-optimal solution.

6. The control method of a refrigerator case production line according to claim 1, characterized by, The target function satisfies: The result of the target function of each candidate solution before each output optimal solution is recorded as the sum of the normalized result of the energy consumption of each candidate solution before each output optimal solution and the negative correlation mapping of the product qualification rate of each candidate solution before each output optimal solution.

7. The control method of a refrigerator case production line according to claim 1 or 6, characterized in that, The energy consumption is the energy consumption in a preset time when the candidate solution is applied to control the refrigerator shell production line.

8. The control method of a refrigerator case production line according to claim 1 or 6, characterized by, The product qualification rate is a product qualification rate in historical refrigerator shell detection data obtained in advance when the refrigerator shell production line is controlled by using the candidate solution.

9. The control method of the refrigerator case production line according to claim 1, characterized by, The screening the optimal solution comprises: Based on the adaptive loudness decay coefficient and the initial loudness of the bat algorithm, the loudness value corresponding to each round of iteration is obtained when the bat algorithm outputs the optimal solution each time; In each round of iteration, in response to the random number generated by the bat algorithm being less than the loudness value and the target function result of the new candidate solution obtained in the current round of iteration being less than the candidate solution when the last round of iteration is completed, the new candidate solution obtained in the current round of iteration is taken as the candidate solution when the current round of iteration is completed, otherwise, the candidate solution when the current round of iteration is completed is the candidate solution when the last round of iteration is completed; The candidate solution when the last round of iteration is completed is taken as the optimal solution output by the bat algorithm each time.

10. Control system based on a refrigerator shell production line, characterized by, Comprise: A processor and a memory, the memory stores computer program instructions, when the computer program instructions are executed by the processor, the control method based on the refrigerator shell production line according to any one of claims 1-9 is realized.

Images