Multi-section intelligent variable-frequency power control system and method for dough kneading machine

Abstract

The invention relates to the field of mechanical automation control, and discloses a multi-section intelligent frequency conversion power control system and method for a dough kneading machine, and the method comprises the steps: carrying out the segmented collection of the power data of the dough kneading machine and the dough state data in the whole dough kneading process according to the process stages, extracting the power characteristics of each stage from the power data of each stage, and obtaining the power characteristics of the dough kneading machine; associating the extracted power characteristics of each stage with dough state data, constructing a power-state mapping table, comparing the power characteristics extracted in each stage in the current dough kneading process with the power characteristics of the same process stage in historical data, calculating a characteristic change rate, and determining the state of the dough according to the characteristic change rate. The characteristic change rate is matched with the corresponding power data in the power-state mapping table to obtain the dough state prediction index, and then the dough state prediction index is compared with the reasonable range of the set dough state prediction index to judge whether frequency conversion adjustment is triggered or not, so that the accuracy and rationality of frequency conversion adjustment are improved; and the stability of the dough kneading process and the consistency of the product quality are ensured.

Description

Technical Field

[0001] This invention relates to the field of mechanical automation control technology, and more specifically to a multi-segment intelligent variable frequency power control system and method for a dough kneading machine. Background Technology

[0002] The existing multi-segment intelligent variable frequency power control system of dough kneading machine mainly uses a variety of sensors to collect key parameters in real time during the kneading process, such as dough temperature, humidity, kneading time, motor current, etc., and feeds this data back to the control system. The control system analyzes and processes the collected data according to the preset algorithm and model, and automatically adjusts the motor power and operating status. However, the existing multi-segment intelligent variable frequency power control of dough kneading machines still has the following drawbacks: Firstly, the existing multi-segment intelligent frequency conversion power control of dough kneading machines often does not divide the kneading process into detailed stages when controlling power, and adopts a uniform or relatively coarse power control strategy, which cannot accurately adjust the power according to the actual needs of the dough at different stages. Secondly, the existing multi-segment intelligent frequency conversion power control of dough kneading machines lacks a clear correlation mechanism between power control and dough state. It is impossible to predict the state of the dough based on the correlation mechanism, and it is impossible to know in advance whether the dough has reached the ideal processing state. Thirdly, the frequency conversion adjustment of the multi-segment intelligent variable frequency power control of existing dough kneading machines may lack scientific judgment standards. Operators often decide whether to make frequency conversion adjustments and the extent of adjustment based on subjective judgment or simple experience rules, which lacks a scientific basis for adjustment. Summary of the Invention

[0003] In order to overcome the above-mentioned defects of the prior art, the present invention provides a multi-segment intelligent variable frequency power control system and method for a dough kneading machine, so as to solve the problems existing in the background art.

[0004] This invention provides the following technical solution: a multi-segment intelligent variable frequency power control system for a dough kneading machine, comprising: Stage-based data acquisition module: Collects the power data of the dough kneading machine in segments according to the process stages throughout the entire kneading process, and simultaneously collects the dough state data of each segment. Then, it preprocesses the power data and dough state data collected in each stage. Power feature extraction module: Extracts power features of each stage from the preprocessed power data of each stage, including time-domain features, frequency-domain features and time-varying features of each stage. Association Analysis Module: This module associates the extracted power characteristics of each stage with the dough state data, establishes the correspondence between power and dough state, analyzes and statistically analyzes the power and state data, and constructs a power-state mapping table based on different power ranges and dough state ranges. Power characteristic change analysis module: It compares the power characteristics extracted at each stage of the current kneading process with the power characteristics of the same process stage in historical data and calculates the characteristic change rate. Feature matching analysis module: Matches the feature change rate with the corresponding power data in the power-state mapping table to obtain the dough state prediction index; Variable frequency trigger module: It compares the dough state prediction index with the set reasonable range of the dough state prediction index to determine whether to trigger variable frequency adjustment.

[0005] Preferably, the segmented acquisition of the power data includes: The entire kneading process is divided into stages according to the process flow. The stages include the initial mixing stage, the gluten development stage, the intensified kneading stage, and the final shaping stage. Power data and dough state data were collected for the initial mixing stage, gluten development stage, strengthening kneading stage, and final shaping stage, and the collected power data and dough state data were classified and stored in the database according to the divided stages. The dough condition data includes temperature, humidity, elasticity, viscosity, and gluten content; The preprocessing steps for the power data and dough state data include: data cleaning, data transformation, and data normalization.

[0006] Preferably, the extraction of the power features includes: The average power data, maximum power data, minimum power data, variance of power data, and ratio of maximum power to effective power of each stage are calculated from the power data of each stage after preprocessing, and are used as the time-domain features of each stage. The power signals in the time domain of each stage are converted into frequency domain signals by Fourier transform to obtain the power spectral density of each stage. The main frequency components of each stage are extracted from the power spectral density of each stage. The frequency domain of each stage is then divided into several frequency bands, and the power energy in each frequency band of each stage is calculated. The power spectral density of each stage, the main frequency components of each stage, and the power energy in each frequency band of each stage are used as the frequency domain features of each stage. The power change rate of each stage is calculated from the power data of each stage after preprocessing. The ratio of the power change to the time interval between adjacent time points in each stage is obtained. The cumulative power over time in each stage is calculated to obtain the cumulative power of each stage. The power change rate and the cumulative power of each stage are used as the time-varying characteristics of each stage.

[0007] Preferably, the construction of the power-state mapping table includes: Pearson correlation coefficient analysis was used to analyze the linear correlation between the extracted power characteristics of each stage and the dough state data. The time-domain characteristics, frequency-domain characteristics and time-varying characteristics of each stage were respectively correlated with the dough state data. The correlation coefficient between each power characteristic and each dough state index of each stage was calculated. The relationship between power characteristics and dough state was displayed by a graphical visualization method. After completing the correlation analysis between power characteristics and dough state at each stage, the power characteristics of all stages are integrated to obtain a power characteristic set. Based on the integrated power characteristic set and the distribution of dough state data, the data is grouped using an equal-width grouping method. For each power characteristic interval, the number of occurrences of different dough state levels in that interval is counted, and then the probability of occurrence of each dough state level in that power characteristic interval is calculated. Design a table where rows represent different combinations of power characteristics and columns represent different dough state indicators and their levels. Based on the data grouping and statistical results, fill in the dough state indicator level and its probability of occurrence for each power characteristic combination into the table, thus forming a power-state mapping table.

[0008] Preferably, the calculation of the characteristic change rate includes: The power data of the same batch as the current kneading process is selected from the historical records of the database. The current kneading power data and the historical kneading power data are aligned according to the process stage. For each stage, the corresponding power feature values ​​in the current kneading power data and the historical kneading power data are extracted to form a one-to-one correspondence. Calculate the rate of change of each characteristic value at each stage of the current kneading process with the characteristic value at the same stage in the historical kneading process.

[0009] Preferably, the analysis of the dough state prediction index includes: The calculated power characteristic change rates are combined according to the preset power characteristic combination rules. Based on the combined power characteristic change rates, the corresponding power range is determined in the pre-constructed power-state mapping table. Combining the ranges to which each power characteristic change rate belongs, the power range region corresponding to the matching power characteristic combination is found in the power-state mapping table. After determining the power range corresponding to the power feature combination, the probability values ​​corresponding to different levels of each dough state index under the power range are extracted from the power-state mapping table. Based on multiple dough state indices and their corresponding probabilities, a weighted summation method is used to calculate the dough state prediction index, thereby predicting the current state of the dough.

[0010] Preferably, the triggering determination for the frequency conversion adjustment includes: By setting a reasonable range for the dough state prediction index, If the calculated dough state prediction index is not within the reasonable range of the set dough state prediction index, frequency conversion adjustment will be triggered. Based on the magnitude and direction of the deviation from the dough condition prediction index, the speed and power output of the dough kneading machine are adjusted to optimize its operating parameters.

[0011] To achieve the above objectives, the present invention provides the following technical solution: a multi-segment intelligent variable frequency power control method for a dough kneading machine. The multi-segment intelligent variable frequency power control system for a dough kneading machine includes the following steps: S1: Collect the power data of the dough kneading machine in segments according to the process stages throughout the entire kneading process, and collect the dough state data of each stage simultaneously. Then, preprocess the power data and dough state data collected in each stage. S2: Extract the power characteristics of each stage from the power data of each stage after preprocessing, including the time domain characteristics, frequency domain characteristics and time-varying characteristics of each stage. S3: Associate the extracted power characteristics of each stage with the dough state data, establish the correspondence between power and dough state, analyze and statistically analyze the power and state data, and construct a power-state mapping table according to different power ranges and dough state ranges. S4: Compare the power characteristics extracted at each stage of the current kneading process with the power characteristics of the same process stage in historical data, and calculate the characteristic change rate. S5: Match the characteristic rate of change with the corresponding power data in the power-state mapping table to obtain the dough state prediction index; S6: Based on the comparison between the dough state prediction index and the set reasonable range of the dough state prediction index, determine whether to trigger frequency conversion adjustment.

[0012] The technical effects and advantages of this invention are as follows: By collecting the power data of the dough kneading machine in segments according to the process stages throughout the entire kneading process, and simultaneously collecting the dough state data of each segment, the power can be controlled separately according to the characteristics of each specific stage, such as the mixing stage, shaping stage, and proofing stage, thereby improving the accuracy and specificity of power control.

[0013] By associating the extracted power characteristics at each stage with dough state data, a correspondence between power and dough state is established, and a power-state mapping table is constructed. Through analysis and statistical analysis of a large amount of power and state data, the possible states of the dough under different power ranges are clarified, providing a scientific basis for power control. Furthermore, by matching and analyzing the characteristic change rate with the corresponding power data in the power-state mapping table, and comprehensively considering the changes in power characteristics and the power-state relationship in historical data, the current state of the dough can be predicted more accurately, providing a reliable basis for subsequent frequency conversion adjustments.

[0014] By comparing the dough state prediction index with the set reasonable range of the dough state prediction index, it is determined whether to trigger frequency conversion adjustment. This allows for a scientific determination of whether power adjustment is needed and the adjustment range based on the actual state of the dough, improving the accuracy and rationality of frequency conversion adjustment and ensuring the stability of the kneading process and the consistency of product quality. Attached Figure Description

[0015] Figure 1 This is a system structure block diagram of the present invention. Figure 2 This is a diagram illustrating the method steps of the present invention. Detailed Implementation

[0016] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings. In addition, the forms of the various structures described in the following embodiments are merely illustrative. The multi-segment intelligent variable frequency power control system and method for a dough kneading machine involved in the present invention are not limited to the structures described in the following embodiments. All other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0017] like Figure 1 The embodiment shown provides a multi-segment intelligent variable frequency power control system for a dough kneading machine, including: Stage-based data acquisition module: Collects the power data of the dough kneading machine in segments according to the process stages throughout the entire kneading process, and simultaneously collects the dough state data of each segment. Then, it preprocesses the power data and dough state data collected in each stage.

[0018] In this embodiment, the segmented acquisition of power data includes: The entire kneading process is divided into stages according to the process flow. The stages include the initial mixing stage, the gluten development stage, the intensified kneading stage, and the final shaping stage. Power data and dough state data were collected for the initial mixing stage, gluten development stage, strengthening kneading stage, and final shaping stage, and the collected power data and dough state data were classified and stored in the database according to the divided stages. The dough condition data includes temperature, humidity, elasticity, viscosity, and gluten content; The preprocessing steps for the power data and dough state data include: data cleaning, data transformation, and data normalization.

[0019] It should be noted that during the kneading process of the dough mixer, power sensors are used to collect power data in real time at each stage. These stages can be further divided into initial mixing, rapid kneading, and slow relaxation. Simultaneously, multiple sensors are used to collect dough state data. A temperature sensor measures the dough temperature, a humidity sensor measures the dough humidity, and an elasticity tester and a viscosity tester are used to measure the dough's elasticity and viscosity, respectively.

[0020] To eliminate differences in units and numerical ranges among different indicators, the preprocessed power data and dough state data were normalized. Common normalization methods include min-max normalization, which linearly maps the data to... Range. For example, for power data, the calculation formula is: ,in, The power data is normalized. This is the raw power data. and These are the minimum and maximum values ​​of the power data, respectively; the dough state data are then normalized using the same method.

[0021] Power feature extraction module: Extracts power features of each stage from the preprocessed power data of each stage, including time-domain features, frequency-domain features and time-varying features of each stage.

[0022] In this embodiment, the extraction of power features includes: The average power data, maximum power data, minimum power data, variance of power data, and ratio of maximum power to effective power of each stage are calculated from the power data of each stage after preprocessing, and are used as the time-domain features of each stage. The power signals in the time domain of each stage are converted into frequency domain signals by Fourier transform to obtain the power spectral density of each stage. The main frequency components of each stage are extracted from the power spectral density of each stage. The frequency domain of each stage is then divided into several frequency bands, and the power energy in each frequency band of each stage is calculated. The power spectral density of each stage, the main frequency components of each stage, and the power energy in each frequency band of each stage are used as the frequency domain features of each stage. The power change rate of each stage is calculated from the power data of each stage after preprocessing. The ratio of the power change to the time interval between adjacent time points in each stage is obtained. The cumulative power over time in each stage is calculated to obtain the cumulative power of each stage. The power change rate and the cumulative power of each stage are used as the time-varying characteristics of each stage.

[0023] Association Analysis Module: This module associates the extracted power characteristics of each stage with the dough state data, establishes a correspondence between power and dough state, analyzes and statistically analyzes the power and state data, and constructs a power-state mapping table based on different power ranges and dough state ranges.

[0024] In this embodiment, the construction of the power-state mapping table includes: Pearson correlation coefficient analysis was used to analyze the linear correlation between the extracted power characteristics of each stage and the dough state data. The time-domain characteristics, frequency-domain characteristics and time-varying characteristics of each stage were respectively correlated with the dough state data. The correlation coefficient between each power characteristic and each dough state index of each stage was calculated. The relationship between power characteristics and dough state was displayed by a graphical visualization method. After completing the correlation analysis between power characteristics and dough state at each stage, the power characteristics of all stages are integrated to obtain a power characteristic set. Based on the integrated power characteristic set and the distribution of dough state data, the data is grouped using an equal-width grouping method. For each power characteristic interval, the number of occurrences of different dough state levels in that interval is counted, and then the probability of occurrence of each dough state level in that power characteristic interval is calculated. Design a table where rows represent different combinations of power characteristics and columns represent different dough state indicators and their levels. Based on the data grouping and statistical results, fill in the dough state indicator level and its probability of occurrence for each power characteristic combination into the table, thus forming a power-state mapping table.

[0025] It should be specifically noted that Pearson correlation coefficient analysis is used to accurately analyze the linear correlation between the aforementioned power characteristics and dough state data. Specifically, for each power characteristic at each stage, the correlation is calculated with each dough state index (including but not limited to temperature, humidity, elasticity, viscosity, and gluten formation). The Pearson correlation coefficient formula is as follows: ,in For power characteristic data points, This represents the average value of the power characteristic data. For dough condition index data points, Let be the average value of the dough condition index data, n be the total number of data points, and the correlation coefficient r range from 1 to 1. The closer the absolute value is to 1, the stronger the correlation.

[0026] Using professional data visualization tools, various graphs are drawn to intuitively show the relationship between power characteristics and dough state. For example, scatter plots are drawn to show the distribution trend between power characteristics and dough state indicators, and the density and distribution direction of the points can be used to determine whether there is a linear or non-linear relationship; heat maps are drawn to show the magnitude of the correlation coefficient between different power characteristics and dough state indicators, and the color intensity intuitively reflects the strength of the correlation.

[0027] After completing the correlation analysis between the power characteristics at each stage and the dough state, the power characteristics of all stages were integrated. During the integration process, the importance of the power characteristics at each stage and their correlation with the dough state were comprehensively considered. Power characteristics that have a significant impact on the dough state were retained, while redundant or extremely low-correlation characteristics were removed, forming a comprehensive set of power characteristics.

[0028] Based on the integrated power feature set and the actual distribution of the dough state data, data partitioning is performed using equal-width grouping or equal-frequency grouping methods. For power features, such as the average power value, if its value range is... When using equal-width grouping, it is divided into k intervals, each interval having a width of [missing value]. Then the grouping interval is When using equal-frequency grouping, ensure that the number of data points in each interval is approximately equal. For dough condition indicators, divide them into different levels according to the actual evaluation criteria; for example, divide elasticity into three levels: "high," "medium," and "low."

[0029] For each power characteristic interval, count the number of times different dough state levels occur within that interval. Here, 'a' represents the power characteristic interval number, and 'b' represents the dough state level number. Then, the probability of each dough state level occurring within that power characteristic interval is calculated. ,in This represents the total number of data points within the power characteristic range.

[0030] Design a table with a defined row and column structure, where each row represents a different combination of power characteristics. This combination can contain different value ranges for multiple power characteristics, such as "average power at..." And the main frequency component is The columns in the table represent different dough state indicators and their levels, such as "High Elasticity," "Medium Elasticity," "Low Elasticity," "Strong Stickiness," "Medium Stickiness," and "Weak Stickiness." Based on the data grouping and statistical results, the dough state indicator level corresponding to each power characteristic combination and its probability or frequency of occurrence are accurately filled into the table. If a dough state level does not appear under a certain power characteristic combination, 0 is entered in the corresponding cell; if it appears, the calculated probability value is entered to ensure the completeness and accuracy of the table data, thus forming a power-state mapping table.

[0031] Power characteristic change analysis module: It compares the power characteristics extracted at each stage of the current kneading process with the power characteristics of the same process stage in historical data and calculates the characteristic change rate.

[0032] In this embodiment, the calculation of the characteristic rate of change includes: The power data of the same batch as the current kneading process is selected from the historical records of the database. The current kneading power data and the historical kneading power data are aligned according to the process stage. For each stage, the corresponding power feature values ​​in the current kneading power data and the historical kneading power data are extracted to form a one-to-one correspondence. Calculate the rate of change of each characteristic value at each stage of the current kneading process with the characteristic value at the same stage in the historical kneading process.

[0033] Feature matching analysis module: Matches the feature change rate with the corresponding power data in the power-state mapping table to obtain the dough state prediction index.

[0034] In this embodiment, the analysis of the dough state prediction index includes: The calculated power characteristic change rates are combined according to the preset power characteristic combination rules. Based on the combined power characteristic change rates, the corresponding power range is determined in the pre-constructed power-state mapping table. Combining the ranges to which each power characteristic change rate belongs, the power range region corresponding to the matching power characteristic combination is found in the power-state mapping table. After determining the power range corresponding to the power feature combination, the probability values ​​corresponding to different levels of each dough state index under the power range are extracted from the power-state mapping table. Based on multiple dough state indices and their corresponding probabilities, a weighted summation method is used to calculate the dough state prediction index, thereby predicting the current state of the dough.

[0035] It needs to be specifically explained that the calculated power characteristic change rates are combined according to a preset power characteristic combination rule. For example, the power average change rate and the power fluctuation change rate are combined into a power characteristic combination. Based on the combined power characteristic change rate, the corresponding power range is determined in the pre-constructed power-state mapping table (i.e., the table designed above). Specifically, for each power characteristic change rate in the power characteristic combination, it is compared with the preset power range boundary value in the power-state mapping table. Taking the power average change rate as an example, if the power-state mapping table divides the power average change rate into low... ,middle ,high Three ranges, when the calculated average power change rate is At that time, directly combine it with , , By comparing these rates, the range of the average power change rate is determined. The same method is used to determine the range of other power characteristic change rates in the power characteristic combination. Combining the ranges of all power characteristic change rates, the power range region corresponding to the matching power characteristic combination is found in the power-state mapping table. For example, if the average power change rate is in the middle range and the power fluctuation change rate is in the high range, then the power range region corresponding to the high power fluctuation change rate within the average power change rate is found in the mapping table.

[0036] Taking into account multiple dough state indicators and their corresponding probabilities, a weighted summation method is used to calculate the dough state prediction index. The specific calculation formula is as follows: ,in Let j be the weight of the influence of the j-th dough state index on the dough state prediction. This represents the probability that the dough's state index is at level b within a defined power range.

[0037] Variable frequency trigger module: It compares the dough state prediction index with the set reasonable range of the dough state prediction index to determine whether to trigger variable frequency adjustment.

[0038] In this embodiment, the triggering judgment for frequency conversion adjustment includes: By setting a reasonable range for the dough state prediction index, If the calculated dough state prediction index is not within the reasonable range of the set dough state prediction index, frequency conversion adjustment will be triggered. Based on the magnitude and direction of the deviation from the dough condition prediction index, the speed and power output of the dough kneading machine are adjusted to optimize its operating parameters.

[0039] like Figure 2 This embodiment provides a multi-segment intelligent variable frequency power control method for a dough kneading machine, including the following steps: S1: Collect the power data of the dough kneading machine in segments according to the process stages throughout the entire kneading process, and collect the dough state data of each stage simultaneously. Then, preprocess the power data and dough state data collected in each stage. S2: Extract the power characteristics of each stage from the power data of each stage after preprocessing, including the time domain characteristics, frequency domain characteristics and time-varying characteristics of each stage. S3: Associate the extracted power characteristics of each stage with the dough state data, establish the correspondence between power and dough state, analyze and statistically analyze the power and state data, and construct a power-state mapping table according to different power ranges and dough state ranges. S4: Compare the power characteristics extracted at each stage of the current kneading process with the power characteristics of the same process stage in historical data, and calculate the characteristic change rate. S5: Match the characteristic rate of change with the corresponding power data in the power-state mapping table to obtain the dough state prediction index; S6: Based on the comparison between the dough state prediction index and the set reasonable range of the dough state prediction index, determine whether to trigger frequency conversion adjustment.

[0040] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

[0041] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A multi-stage intelligent variable frequency power control system for a dough mixer, characterized in that, The method comprises the following steps: The stage division acquisition module: the power data of the whole kneading process is segmented and acquired according to the process stage, and the dough state data of each stage is synchronously acquired, and then the power data and the dough state data of each stage are preprocessed; The power feature extraction module: the power features of each stage are extracted from the preprocessed power data of each stage, including the time domain features of each stage, the frequency domain features of each stage and the time-varying features of each stage; The correlation analysis module: the extracted power features of each stage are correlated with the dough state data, the corresponding relationship between the power and the dough state is established, the power and the state data are analyzed and counted, and the power-state mapping table is constructed according to different power ranges and dough state ranges; The power feature change analysis module: the power features of each stage in the current kneading process are compared with the power features of the same process stage in the historical data, and the feature change rate is calculated; The feature matching analysis module: the feature change rate is matched with the corresponding power data in the power-state mapping table to obtain the dough state prediction index; The frequency conversion triggering module: the dough state prediction index is compared with the reasonable range of the set dough state prediction index to determine whether the frequency conversion adjustment is triggered.

2. A multi-stage intelligent variable frequency power control system for a dough kneader as claimed in claim 1, wherein, The segmented acquisition of the power data comprises: The whole kneading process is divided into stages according to the process flow, and the divided stages include the initial mixing stage, the gluten formation stage, the intensive kneading stage and the finishing shaping stage; The power data and the dough state data of the initial mixing stage, the gluten formation stage, the intensive kneading stage and the finishing shaping stage are respectively acquired, and the acquired power data and dough state data are classified and stored in the database according to the divided stages; The dough state data includes temperature, humidity, elasticity, viscosity and gluten degree; The preprocessing operation process of the power data and the dough state data includes data cleaning, data conversion and data normalization processing.

3. A multi-stage intelligent variable frequency power control system for a dough kneader as claimed in claim 2, wherein, The extraction of the power features comprises: The average value of the power data in each stage, the maximum value of the power data in each stage, the minimum value of the power data in each stage, the variance of the power data in each stage and the ratio of the maximum power to the effective value power in each stage are calculated from the preprocessed power data of each stage as the time domain features of each stage; The time domain power signals of each stage are converted into frequency domain signals by Fourier transform to obtain the power spectral density of each stage, and the main frequency components of each stage are extracted from the power spectral density of each stage, and then the frequency domain of each stage is divided into several frequency bands, and the power energy in each frequency band of each stage is calculated, and the power spectral density of each stage, the main frequency components of each stage and the power energy in each frequency band of each stage are taken as the frequency domain features of each stage; The ratio of the power change amount to the time interval of adjacent time points in each stage is calculated from the preprocessed power data of each stage to obtain the power change rate of each stage, and the cumulative value of the power with time in each stage is calculated to obtain the power accumulation of each stage, and the power change rate of each stage and the power accumulation of each stage are taken as the time-varying features of each stage.

4. A multi-stage intelligent variable frequency power control system for a dough kneader as claimed in claim 3, wherein, The construction of the power-state mapping table includes: The linear correlation between the extracted power features of each stage and the dough state data is analyzed by using Pearson correlation coefficient analysis method. The time domain features of each stage, the frequency domain features of each stage, and the time-varying features of each stage are respectively correlated with the dough state data. The correlation coefficients between each power feature of each stage and each dough state index are calculated. The relationship between the power features and the dough state is displayed by a visual method combining graphics. After the correlation analysis of the power features of each stage and the dough state, the power features of all stages are integrated to obtain a power feature set. According to the distribution of the integrated power feature set and the dough state data, the data is grouped by using the equal-width grouping method. The occurrence frequency of different dough state levels in each power feature interval is counted, and the occurrence probability of each dough state level in the power feature interval is calculated. A table is designed, with rows representing different power feature combinations and columns representing different dough state indexes and their levels. According to the results of data grouping and statistics, the dough state index levels corresponding to each power feature combination and their occurrence probabilities are filled into the table, thereby constructing a power-state mapping table.

5. A multi-stage intelligent variable frequency power control system for a dough kneader as claimed in claim 4, wherein, The calculation of the feature change rate includes: From the historical records of the database, the power data of the same batch as the current dough mixing process is selected. The current dough mixing power data and the historical dough mixing power data are aligned according to the process stages. For each stage, the corresponding power feature values of the current dough mixing power data and the historical dough mixing power data are extracted to form a one-to-one correspondence. The feature change rate between each feature value of the current dough mixing process and the same stage feature value of the historical dough mixing process is calculated.

6. A multi-stage intelligent variable frequency power control system for a dough mixer as claimed in claim 5, wherein, The analysis of the dough state prediction index includes: The calculated power feature change rates are combined according to the preset power feature combination rules. Based on the combined power feature change rates, the corresponding power range is determined in the pre-constructed power-state mapping table. The ranges of the power feature change rates are integrated to find the matching power range area corresponding to the power feature combination in the power-state mapping table. After determining the power range corresponding to the power feature combination, the probability values corresponding to different levels of each dough state index in the power range are extracted from the power-state mapping table. According to the multiple dough state indexes and their corresponding probabilities, the weighted sum method is used to calculate the dough state prediction index, thereby predicting the current state of the dough.

7. A multi-stage intelligent variable frequency power control system for a dough mixer as claimed in claim 6, wherein, The trigger judgment of the frequency conversion adjustment includes: By setting a reasonable range of the dough state prediction index as , if the calculated dough state prediction index is not within the set reasonable range of the dough state prediction index, the frequency conversion adjustment is triggered; According to the size and direction of the deviation range of the dough state prediction index, the speed and power output of the dough mixer are adjusted to optimize the operation parameters of the dough mixer.

8. A multi-stage intelligent variable frequency power control method for dough kneader, using a multi-stage intelligent variable frequency power control system for dough kneader as claimed in any one of claims 1-7, characterized in that, The method includes the following steps: S1: The power data of the dough mixer in the entire dough mixing process is segmented according to the process stages, and the dough state data of each stage is collected synchronously. The collected power data and dough state data of each stage are preprocessed. S2: extracting power features of each stage from the pretreated power data of each stage, including time domain features of each stage, frequency domain features of each stage and time-varying features of each stage; S3: associating the extracted power features of each stage with the dough state data, establishing the corresponding relationship between power and dough state, analyzing and counting the power and state data, and constructing a power-state mapping table according to different power ranges and dough state ranges; S4: comparing the extracted power features of each stage in the current dough kneading process with the power features of the same process stage in the historical data, and calculating the feature change rate; S5: matching and analyzing the feature change rate with the corresponding power data in the power-state mapping table to obtain a dough state prediction index; S6: comparing the dough state prediction index with a reasonable range of the set dough state prediction index to determine whether to trigger frequency conversion adjustment.

Images