Homogenizing apparatus for solid powders and method of mixing control thereof
By using real-time acquisition and multi-dimensional signal analysis, the homogeneous mixing state of solid powder is evaluated, which solves the problem of insufficient accuracy in the evaluation of mixing state in traditional DCS and realizes efficient homogenization processing and control of solid powder.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- HUNAN JINHEYI MINING TECH CO LTD
- Filing Date
- 2026-02-11
- Publication Date
- 2026-05-15
AI Technical Summary
In the homogenization and mixing control of solid powders, the existing technology of DCS relies on single-dimensional sensor data, which leads to insufficient accuracy in assessing the homogenization and mixing state. This affects the adjustment accuracy of the DCS's built-in model predictive control algorithm, resulting in poor homogenization treatment of solid powders.
By acquiring the torque voltage signal of the stirring shaft and the vibration signal of the inner wall of the mixing chamber in real time, and combining multi-dimensional signal analysis, the homogeneous mixing state of the solid powder is evaluated, and the softening coefficient in the model predictive control algorithm is adjusted to achieve precise control.
It improves the homogenization effect of solid powders, enhances the robustness of homogenization equipment, and ensures the accuracy and stability of the mixing process.
Smart Images
Figure CN121669075B_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of homogenization and mixing control technology for solid powders, specifically to a homogenization device for solid powders and a mixing control method thereof. Background Technology
[0002] As global mineral resource extraction shifts towards deeper and more complex conditions, mine backfilling technology, as a core component of "green mining" and "zero-waste mining," can not only treat goaf areas, control ground pressure, and prevent surface subsidence, but also dispose of industrial solid waste such as tailings and coal gangue, achieving resource recycling. However, the performance of new mine backfilling materials depends on the thorough homogeneous mixing of powders at both the macroscopic and microscopic scales.
[0003] Distributed Control Systems (DCS), as an automated system for industrial process control, offer advantages in decentralized control and real-time monitoring. However, in the homogenization and mixing control of solid powders, traditional DCS relies on single-dimensional sensor data, which may result in insufficient accuracy in assessing the homogenization and mixing state of the solid powders. This reduces the adjustment accuracy of the softening coefficient in the DCS's built-in model predictive control algorithm, thereby affecting the homogenization effect of the solid powders and potentially leading to performance issues when using the homogenized solid powders as new mine backfill materials. Summary of the Invention
[0004] In view of the above, it is necessary to provide a homogenizing device for solid powders and a mixing control method thereof, which improves the robustness of the homogenizing device and thus enhances the homogenization effect of solid powders compared with traditional homogenizing devices and mixing control methods.
[0005] In a first aspect, embodiments of this application provide a mixing control method for a homogenizing device for solid powders, the method comprising the following steps:
[0006] The torque voltage signal of the mixing shaft of the mixer in the solid powder homogenizing equipment is collected in real time, as well as the vibration signals of different directions at preset positions on the inner wall of the mixing chamber.
[0007] The evaluation cycle of the homogeneous mixing state of the solid powder in the mixing chamber is preset; the torque voltage fluctuation of each cycle is obtained by the amplitude dispersion and rise of the torque voltage signal in each cycle, and the low frequency energy anomaly of each cycle is obtained by combining the periodic decline of the low frequency component of the torque voltage signal in each cycle and the energy ratio of the low frequency component, and the torque anomaly of each cycle is obtained.
[0008] By measuring the differences in vibration signals in each direction at different heights within each cycle, the vibration differences in each direction within each cycle are obtained. Then, by combining the distribution of peak values in the vibration signals at all positions within each cycle in each direction, the vibration anomaly degree in each direction within each cycle is obtained. This is then fused with the torque anomaly degree to evaluate the homogeneous mixing state of the solid powder within each cycle. This allows for the adjustment of the softening coefficient in the model predictive control algorithm, thereby controlling the mixing state of the solid powder.
[0009] In one embodiment, the process of obtaining the torque voltage fluctuation is as follows:
[0010] Calculate the dispersion of the amplitude of the torque voltage signal within each cycle;
[0011] Each cycle is divided into windows. By analyzing the changes in torque voltage signal between any two adjacent windows within each cycle, the degree of rise of torque voltage signal within each cycle is measured.
[0012] In one embodiment, the process of obtaining the low-frequency energy anomaly is as follows:
[0013] Obtain the low-frequency subband of the torque voltage signal within each window, and the periodic intensity of the energy values at all frequencies in the low-frequency subband;
[0014] The difference in periodic intensity between each window and its adjacent next window within each period is denoted as the periodic difference.
[0015] Calculate the cumulative energy value of the torque voltage signal at all frequencies in the low-frequency sub-band within each window; calculate the sum of the cumulative energy values of the torque voltage signal at all frequencies within each window; calculate the ratio of the cumulative value to the sum.
[0016] The low-frequency energy anomaly is positively correlated with all the periodic differences in each period and also positively correlated with the sum of all the ratios in each period.
[0017] In one embodiment, the torque anomaly is the normalized result of the product of the torque voltage fluctuation and the low-frequency energy anomaly.
[0018] In one embodiment, the vibration difference is calculated as follows:
[0019] The positions are evenly distributed on the inner wall of the mixing chamber according to a pre-set row and column rule;
[0020] The difference in vibration signal between any two adjacent rows and each column position in each direction is measured separately. The cumulative value of the measurement results of the difference in vibration signal between any two adjacent rows and each column position in each direction in each period is taken as the vibration difference in each direction in each period.
[0021] In one embodiment, the vibration anomaly degree is calculated as follows:
[0022] Calculate the interquartile range of all maxima in time for each vibration signal; count the total number of maxima for each vibration signal.
[0023] For each direction, calculate the sum of the products of the normalized interquartile range of the vibration signals at all locations and the total number;
[0024] The vibration anomaly degree is the normalized value of the product of the vibration difference and the sum.
[0025] In one embodiment, the process of evaluating the homogeneous mixing state of the solid powder in each period is as follows:
[0026] Based on the torque anomaly and the vibration anomaly in all directions within each cycle, a multi-attribute decision method is used to obtain the comprehensive anomaly coefficient for each cycle, which is used to evaluate the homogeneous mixing state of the solid powder within each cycle.
[0027] In one embodiment, the method for evaluating the homogeneous mixing state of the solid powder in each period is as follows:
[0028] Obtain the segmentation threshold of the comprehensive anomaly coefficient of a preset number of homogenized devices over multiple historical periods;
[0029] If the overall anomaly coefficient of each cycle is less than the segmentation threshold, it is determined that the solid powder is homogeneously mixed in each cycle; otherwise, it is determined that the solid powder is not homogeneously mixed in each cycle.
[0030] In one embodiment, the method for adjusting the softening coefficient in the model predictive control algorithm is as follows:
[0031] When the comprehensive anomaly coefficient of each period is greater than or equal to the segmentation threshold, the difference between 1 and the comprehensive anomaly coefficient of each period is recorded as the adjustment difference. The product of the preset initial value of the softening coefficient in the model predictive control algorithm and the adjustment difference is used as the softening coefficient in the model predictive control algorithm in the next period; otherwise, the softening coefficient in the model predictive control algorithm is not adjusted.
[0032] Secondly, embodiments of this application also provide a homogenizing device for solid powder, wherein the device contains:
[0033] The process monitoring module is used to collect the torque voltage signal of the mixing shaft of the mixer of the solid powder homogenizer in real time, as well as the vibration signals of different directions at preset positions on the inner wall of the mixing chamber.
[0034] The data analysis module is used to preset the evaluation cycle of the homogeneous mixing state of solid powder in the mixing chamber; by measuring the amplitude dispersion and rise of the torque voltage signal in each cycle, the torque voltage fluctuation of each cycle is obtained; and by combining the periodic decline of the low-frequency component of the torque voltage signal in each cycle and the energy ratio of the low-frequency component, the low-frequency energy anomaly of each cycle is obtained, and the torque anomaly of each cycle is obtained.
[0035] By measuring the differences in vibration signals in each direction at different heights within each cycle, the vibration differences in each direction within each cycle are obtained. The distribution of peak values in the vibration signals at all positions within each cycle in each direction is combined to obtain the vibration anomaly degree in each direction within each cycle. This is then fused with the torque anomaly degree to evaluate the homogeneous mixing state of the solid powder within each cycle.
[0036] The DCS control module is used to adjust the softening coefficient in the model predictive control algorithm, thereby controlling the mixing state of solid powder.
[0037] This application has at least the following beneficial effects:
[0038] This application can quantify and analyze the characteristics of torque variation during the mixing process by calculating torque voltage fluctuation and low-frequency energy anomaly. Torque voltage fluctuation reflects the degree of change in the torque signal, while low-frequency energy anomaly reveals the variation law of low-frequency components in the torque signal. By combining torque voltage fluctuation and low-frequency energy anomaly, the torque anomaly is obtained, which can more accurately characterize the degree of torque anomaly and reflect the mixing situation of solid powder in the mixing chamber from different perspectives.
[0039] Furthermore, by comprehensively considering the differences in vibration signals and peak distribution in each direction at different height positions, the convective and shear mixing states of solid powder in the mixing chamber can be fully evaluated. By integrating multi-dimensional vibration anomaly and torque anomaly, the homogeneous mixing state of solid powder can be evaluated more comprehensively and accurately, avoiding the one-sidedness of evaluation by a single index.
[0040] Furthermore, based on the homogenized mixing state of the solid powder obtained from the evaluation, the softening coefficient in the model predictive control algorithm is adjusted, thereby controlling the power of the stirring motor. This enables precise control of the mixing state of the solid powder, thereby improving the mixing effect. This dynamic adjustment mechanism based on the evaluation of the mixing state allows the homogenizing equipment to automatically adjust the control strategy according to the real-time changes in the mixing state, quickly respond to and optimize the mixing process, improve the robustness of the homogenizing equipment, and thus enhance the homogenization effect of the solid powder. Attached Figure Description
[0041] To more clearly illustrate the technical solutions and advantages in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0042] Figure 1 A block diagram of a homogenizing device for solid powder provided in one embodiment of this application;
[0043] Figure 2 A flowchart illustrating the steps of a mixing control method for a homogenizing device for solid powder, provided in one embodiment of this application;
[0044] Figure 3 This is a schematic diagram illustrating the process of adjusting the softening factor. Detailed Implementation
[0045] In the description of the embodiments in this application, the words "exemplary," "or," and "for example" are used to indicate examples, illustrations, or descriptions. Any embodiment or design scheme described as "exemplary" or "for example" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or design schemes. Specifically, the use of the words "exemplary," "or," and "for example" is intended to present the relevant concepts in a specific manner.
[0046] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. It should be understood that, unless otherwise stated, " / " in this application means "or".
[0047] It should also be noted that the terms "first" and "second" in this application are used to distinguish similar objects, rather than to describe a specific order or sequence.
[0048] The components of a homogenizing device for solid powder provided in this application are described in detail below with reference to the accompanying drawings.
[0049] Please see Figure 1 The diagram shows a block diagram of a homogenizing device for solid powder according to an embodiment of this application. The homogenizing device includes: a support frame, a mixing chamber, a mixer, a pneumatic system, an enhanced mixing unit, a process monitoring module, a data analysis module, a DCS control module, and a human-machine interaction module.
[0050] The specific details of each component of the homogenizing equipment are as follows:
[0051] (1) Bearing frame: including the frame body and the chute, and coaxially distributed with the mixing chamber, supporting the parallel expansion of multiple chambers;
[0052] (2) Mixing chamber: A vertical, sealed chamber is adopted, including an inlet and an outlet, with the axis distributed vertically, and the upper end being... It has two feed inlets and one discharge outlet at the bottom, which are coaxially nested with the support frame;
[0053] (3) Mixer: including feeding hood, recovery hood, telescopic protective sleeve, lifting drive mechanism, stirring motor, stirring shaft, stirring blade, and aeration disc. The feeding hood and recovery hood are connected by the lifting drive mechanism to form an adjustable sealed working chamber. Several through holes are arranged on the lower end face of the feeding hood and the upper end face of the recovery hood. The through holes are evenly distributed around the axis of the mixing chamber and the diameter of the through holes is 2mm. A guide cavity is set in both the feeding hood and the recovery hood. The guide cavity is spiral.
[0054] (4) Pneumatic system: including booster pump, delivery pipe, return pipe, guide pipe and control valve. The feed port is connected to the booster pump through the delivery pipe, and the discharge port is connected to the booster pump through the return pipe. The booster pump connected to the delivery pipe is connected to the upper end face of the mixing chamber, and the booster pump connected to the return pipe is connected to the lower end face of the mixing chamber. The dual pump design forms a negative pressure environment, which facilitates independent control of the air pressure in the pipe. The pneumatic system, as an auxiliary unit, maintains constant negative pressure suction and periodic pulse aeration during the homogenization and mixing of solid powder. Its control logic is independent of the model predictive control loop of the stirring motor, which aims to provide a flow state for solid powder during the homogenization and mixing process.
[0055] (5) Enhanced mixing unit: including conveying auger and vibration mechanism, with auger built into the guide pipe and vibration mechanism set on the wall of the mixer, with at least one in each of the upper and lower halves;
[0056] (6) Process monitoring module, used to collect in real time the torque voltage signal of the mixing shaft of the mixer of the solid powder homogenizer, as well as the vibration signals of different directions at preset positions on the inner wall of the mixing chamber;
[0057] (7) Data analysis module, used to preset the evaluation cycle of the homogeneous mixing state of solid powder in the mixing chamber; by the amplitude dispersion and rise of the torque voltage signal in each cycle, the torque voltage fluctuation of each cycle is obtained, and by combining the periodic decline of the low frequency component of the torque voltage signal in each cycle and the energy ratio of the low frequency component, the low frequency energy anomaly of each cycle is obtained, and the torque anomaly of each cycle is obtained.
[0058] By measuring the differences in vibration signals in each direction at different heights within each cycle, the vibration differences in each direction within each cycle are obtained. The distribution of peak values in the vibration signals at all positions within each cycle in each direction is combined to obtain the vibration anomaly degree in each direction within each cycle. This is then fused with the torque anomaly degree to evaluate the homogeneous mixing state of the solid powder within each cycle.
[0059] (8) DCS control module, used to adjust the softening coefficient in the model predictive control algorithm, thereby controlling the mixing state of solid powder;
[0060] (9) Human-computer interaction module, used to display various parameters in the homogenization and mixing process of solid powder.
[0061] The following description, in conjunction with the accompanying drawings, details a specific scheme for a homogenizing device for solid powder and its mixing control method provided in this application.
[0062] Please see Figure 2 The diagram illustrates a flowchart of a mixing control method for a homogenizing device for solid powder according to an embodiment of this application. The method includes the following steps:
[0063] Step 1: Real-time acquisition of torque voltage signal of the mixing shaft of the mixer in the solid powder homogenizer, as well as vibration signals at preset positions and directions on the inner wall of the mixing chamber.
[0064] A torque sensor is deployed on the mixing shaft of the mixer in the solid powder homogenizer to collect the torque voltage signal of the mixing shaft in real time. Multiple triaxial accelerometers are deployed on the inner wall of the mixing chamber of the homogenizer to collect the triaxial vibration signal at the installation position of the triaxial accelerometer in real time.
[0065] The triaxial accelerometers are deployed as follows: the mixing chamber is evenly divided into N layers along the vertical axis, and M triaxial accelerometers are evenly deployed on the inner wall of the mixing chamber at the center height of each layer. That is, the triaxial accelerometers are arranged in rows at the same height on the inner wall of the mixing chamber, and in columns at different heights. When the Distributed Control System (DCS) starts, it performs zero-point calibration on the torque sensor and performs a self-test on the triaxial accelerometers. Since the solid powder is mainly affected by convective mixing and shear mixing forces during homogeneous mixing, this application only analyzes the lateral vibration signals and axial vibration signals acquired by the triaxial accelerometers.
[0066] In this embodiment, the acquisition frequency of both the torque sensor and the triaxial accelerometer is set to 1KHz. The acquisition frequency value is preset by the user and can be set by the implementer according to the actual situation. This application does not impose any special restrictions.
[0067] In this embodiment, the values of N and M are 3 and 4, respectively. The values of N and M are preset by the user and can be set by the implementer according to the actual situation. This application does not impose any special restrictions.
[0068] Torque-voltage signals, lateral vibration signals, and axial vibration signals are synchronously read using a high-speed acquisition card and written to the circular buffer of the DCS server. The timestamps of the torque-voltage signals, lateral vibration signals, and axial vibration signals are unified to ensure the accuracy of subsequent time-series analysis. The sampled values of the torque-voltage signals, lateral vibration signals, and axial vibration signals are normalized to eliminate the influence of different dimensions on the subsequent assessment of the solid powder mixing state. Furthermore, to avoid interference from external environmental noise, the normalized torque-voltage signals, lateral vibration signals, and axial vibration signals are denoised. The denoised torque-voltage signals, lateral vibration signals, and axial vibration signals are then pushed to the historical database of the DCS master station via Modbus TCP for storage.
[0069] In this embodiment, the method for normalizing the sampled values in the torque voltage signal, lateral vibration signal, and axial vibration signal is as follows: the sampled values in the torque voltage signal, lateral vibration signal, and axial vibration signal are mapped to the interval [0,1] by using the Z-Score normalization method and the maximum value normalization method respectively. The Z-Score normalization method is used first, and then the maximum value normalization method is used. The Z-Score normalization method and the maximum value normalization method are well known techniques and will not be described in detail in this application.
[0070] In this embodiment, wavelet transform algorithm is used to denoise the normalized torque voltage signal, lateral vibration signal, and axial vibration signal respectively. Wavelet transform algorithm is a known technology and will not be described in detail in this application. As other implementation methods, based on the ability to denoise the normalized torque voltage signal, lateral vibration signal, and axial vibration signal respectively, implementers may use other existing feasible technologies, and this application does not impose any special restrictions.
[0071] Step 2: Preset the evaluation cycle of the homogeneous mixing state of the solid powder in the mixing chamber; obtain the vibration anomaly degree in each direction within each cycle; evaluate the homogeneous mixing state of the solid powder within each cycle.
[0072] Traditional DCS relies on built-in model predictive control algorithms to control the homogenization of solid powders. The softening coefficient in the model predictive control algorithm directly affects the response speed and smoothness of the homogenization process. It needs to be adjusted in real time according to the actual homogenization state of the solid powders. Therefore, it is necessary to first evaluate the homogenization state of the solid powders in the mixing chamber.
[0073] In the homogeneous mixing process of solid powders, the shear resistance generated by the powder during stirring is directly converted into torque on the stirring shaft. Torque is a direct measure of the energy input of the mixer, and the torque voltage signal is closely related to the flow structure of the solid powder within the mixer. Specifically, the worse the homogeneous mixing quality of the solid powder in the mixing chamber, the more uneven the local shear strength will be caused by the dead zones or backflow zones formed by the powder in the mixing chamber, resulting in instantaneous changes in torque and a corresponding increase in the fluctuation amplitude of the torque voltage signal. Furthermore, when solid powders agglomerate, the impact of the stirring blades will generate instantaneous high torque, further exacerbating the increase in the torque voltage signal.
[0074] Based on the above analysis, this application pre-determines the evaluation cycle for the homogeneous mixing state of solid powder in the mixing chamber.
[0075] In this embodiment, the cycle length is 20 minutes. The cycle length is preset by the user and can be set by the implementer according to the actual situation. This application does not impose any special restrictions.
[0076] Taking the i-th cycle as an example, the torque voltage fluctuation in the i-th cycle is obtained by analyzing the amplitude dispersion and rise of the torque voltage signal within the i-th cycle. Specifically:
[0077] Calculate the dispersion of the amplitude of the torque voltage signal within the i-th cycle;
[0078] The i-th period is divided into windows. By analyzing the changes in the torque voltage signal between any two adjacent windows in the i-th period, the degree of rise of the torque voltage signal in the i-th period is measured.
[0079] The torque voltage fluctuation in the i-th cycle is positively correlated with the results of the dispersion and the rise rate measurements, respectively.
[0080] It should be noted that: dispersion refers to the degree of unevenness in the distribution of data, which can be achieved by calculating the coefficient of variation, information entropy, variance, etc. This application does not impose any special restrictions on this.
[0081] It should be noted that positive correlation means that the variables change in the same direction; when one variable increases, the other variable also increases, and when one variable decreases, the other variable also decreases.
[0082] In this embodiment, the window length is 10 seconds. The window length is preset by the user and can be set by the implementer according to the actual situation. This application does not impose any special restrictions.
[0083] In this embodiment, the dispersion is the coefficient of variation. The formula for measuring the rise of the torque voltage signal in the i-th period is: In the formula, This represents the rise metric of the torque voltage signal during the i-th cycle. This represents the total number of windows within the i-th period; , represents the normalized slope of the fitted straight line of the torque voltage signal within the j-th and (j-1)-th windows in the i-th period, respectively; exp() represents the exponential function with the natural constant as the base, used to map the data to positive numbers; norm() represents the normalization function, which in this embodiment is obtained using the Min-Max normalization method. The normalized value is calculated as follows: The larger the calculated rise metric value, the higher the rise of the torque voltage signal in the i-th period. The product of the rise metric value of the torque voltage signal in the i-th period and the dispersion is taken as the torque voltage fluctuation in the i-th period. The least squares method is used to obtain the fitted straight line of the torque voltage signal, and the Min-Max normalization method is used to obtain the normalized value of the slope. Both the least squares method and the Min-Max normalization method are well-known techniques and will not be elaborated upon in this application. As other implementation methods, based on the ability to obtain the fitted straight line of the torque voltage signal, the implementer can use other existing feasible techniques, such as linear regression analysis, weighted least squares, etc., and this application does not impose any special limitations.
[0084] It should be added that if the dispersion is 0, in order to avoid the calculation result of torque voltage fluctuation being forced to 0, the dispersion needs to be mapped to a positive number before subsequent calculations. There are many ways to map the data to a positive number, and the implementer can choose other existing feasible methods. In this embodiment, the dispersion is used as the exponent of an exponential function with the natural constant as the base to achieve the purpose of mapping the dispersion to a positive number.
[0085] It should be noted that time-series differential analysis is a commonly used method for evaluating signal changes in the field of signal analysis. This application reflects the degree of rise of the torque voltage signal by analyzing the changes in the torque voltage signal within adjacent short time intervals. For example, by calculating the difference in slope, the larger the difference, the stronger the instantaneous impact of powder agglomerates on the stirring blades in the homogenizing equipment, the more severe the instantaneous torque, the greater the short-term torque energy surge, and the higher the degree of rise of the torque voltage signal. In the homogenization and mixing process of solid powder, the worse the homogenization and mixing quality of the powder in the mixing chamber, the greater the instability of the torque voltage signal amplitude, the more obvious the rise of the torque voltage signal, and the greater the calculated torque voltage fluctuation.
[0086] Furthermore, when the homogeneous mixing quality of the solid powder in the mixing chamber is poor, significant large-scale structural flow phenomena, such as backflow and accumulation, will occur within the mixing chamber. These phenomena will lead to an increase in the energy proportion of the low-frequency components in the torque voltage signal, and the periodic decrease of the low-frequency components will be more pronounced.
[0087] Based on the above analysis, the low-frequency energy anomaly degree of the i-th cycle is obtained by observing the periodic decrease of the low-frequency component of the torque voltage signal within the i-th cycle and the energy proportion of the low-frequency component. Specifically:
[0088] Obtain the low-frequency and high-frequency sub-bands of the torque voltage signal within each window, as well as the periodic intensity of the energy values at all frequencies in the low-frequency sub-band; the greater the periodic intensity, the more regular the periodicity of the low-frequency components of the torque voltage signal.
[0089] In this embodiment, the expression for the low-frequency energy anomaly degree of the i-th cycle is:
[0090] In the formula, represents the low-frequency energy anomaly of the i-th cycle; exp() represents an exponential function with the natural constant as the base; The difference in periodic intensity between the j-th window and its adjacent next window within the i-th period is denoted as the periodic difference. This represents the total number of windows within the i-th period; This represents the cumulative energy value of the torque voltage signal at all frequencies in the low-frequency subband within the j-th window of the i-th period; The function represents the sum of the energy values of the torque voltage signal at all frequencies within the j-th window of the i-th period; norm() represents the normalization function, which in this embodiment is obtained using the Min-Max normalization method. The normalized value.
[0091] In this embodiment, the method for obtaining the low-frequency sub-band and the high-frequency sub-band is as follows: the decomposition level of the wavelet transform algorithm is set to 1 level, the wavelet basis is selected as db4, the low-frequency sub-band and the high-frequency sub-band are obtained using the wavelet transform algorithm, and the seasonal intensity of the energy value of the torque voltage signal at all frequencies in the low-frequency sub-band within each window is obtained using the STL (Seasons on the Land of the Tide) algorithm, and used as the periodic intensity. Among them, the wavelet transform algorithm and the STL algorithm are well-known technologies, and will not be described in detail in this application. As other implementation methods, based on the ability to obtain the low-frequency sub-band and the periodic intensity, the implementer can use other existing feasible technologies, and this application does not impose any special restrictions.
[0092] It should be noted that in the field of signal analysis, differential accumulation is a commonly used method for evaluating the cumulative effect of trends. In order to measure the periodic intensity decay of the torque voltage signal in the low-frequency subband during the i-th period, this application obtains the cumulative reduction in periodic intensity through differential accumulation. It reflects the energy proportion of the low-frequency component of the torque voltage signal within a short time interval; the greater the calculated low-frequency energy anomaly, the more the energy proportion of the low-frequency component of the torque voltage signal increases in the i-th period, and the more obvious the periodic decrease of the low-frequency component is.
[0093] Furthermore, the torque anomaly degree in the i-th cycle is obtained by combining the torque voltage fluctuation degree and the low-frequency energy anomaly degree, and the expression is as follows:
[0094] In the formula, Represents the torque anomaly in the i-th cycle; norm() represents the normalization function, used to normalize the torque. Mapped to the interval [0,1]; This represents the torque voltage fluctuation during the i-th cycle; This represents the low-frequency energy anomaly degree in the i-th period. In this embodiment, the Min-Max normalization method is used to obtain... The normalized value.
[0095] It should be noted that the torque anomaly degree is obtained by combining the evaluation results from multiple dimensions; the larger the calculated torque anomaly degree, the worse the homogeneous mixing quality.
[0096] In the mixing control process of homogenizing equipment, relying solely on torque anomaly to assess the homogenization quality of solid powder still has certain drawbacks. Specifically, it does not consider the variation characteristics of the mixing chamber in the dimension of vibration signals and lacks analysis of the convective mixing and shear mixing state of solid powder at different height levels in the mixing chamber. This may result in inaccurate assessment of the homogenization mixing state of solid powder, thus failing to provide precise mixing control commands and leading to poor performance of solid powder.
[0097] Specifically, during the homogeneous mixing of solid powders, insufficient uniformity of convective mixing forces within the mixing chamber leads to a decrease in powder flowability. This insufficient flowability results in poorer convective mixing, causing the powder's friction mode to shift from rolling friction to sliding friction, triggering inelastic collisions and transient impacts. Ultimately, this results in a severely irregular multi-peak distribution of the lateral vibration signal. Furthermore, the formation of accumulation and dead zones during mixing further exacerbates the differences in lateral vibration signals at different height levels.
[0098] Based on the above analysis, the differences in lateral vibration in each cycle are obtained by analyzing the differences in lateral vibration signals at different heights within each cycle. Specifically:
[0099] Extract the features from each transverse vibration signal, calculate the feature differences of the transverse vibration signals at each column position between any two adjacent rows, and take the cumulative value of the feature differences at all column positions between any two adjacent rows in the i-th period as the transverse vibration difference in the i-th period.
[0100] In this embodiment, the expression for the lateral vibration difference in the i-th period is:
[0101] In the formula, The value represents the lateral vibration difference in the i-th cycle; N represents the number of rows where the triaxial accelerometers are installed on the inner wall of the mixing chamber; M represents the number of columns where the triaxial accelerometers are installed on the inner wall of the mixing chamber. , These represent the time-series maxima of the transverse vibration signal at the nth row and the mth column of the (n+1)th row within the i-th period; DTW() represents the Dynamic Time Warping (DTW) distance, which is a well-known technique and will not be elaborated upon here. The time-series maxima of the transverse vibration signal are characteristic features of the transverse vibration signal.
[0102] In another embodiment, the calculation process for the lateral vibration difference in the i-th period is as follows:
[0103] The energy values of each transverse vibration signal at all frequencies in the high-frequency subband are arranged in ascending order to form a high-frequency energy sequence for each transverse vibration signal. The difference in the high-frequency energy sequence between any two rows and columns is measured. The cumulative value of the difference in the high-frequency energy sequence between any two rows and columns is taken as the transverse vibration difference of the i-th cycle. The greater the difference in the high-frequency energy of the transverse vibration signals at different height levels, the more severe the inelastic collisions and transient impacts caused by the decrease in powder flowability during homogeneous mixing, and the worse the convective mixing uniformity of the solid powder.
[0104] Among them, the difference measurement results between high-frequency energy sequences are obtained by using DTW distance. As another implementation method, based on the ability to measure the difference between high-frequency energy sequences, the implementer may adopt other existing feasible technologies, such as JS divergence, etc. This application does not impose any special restrictions.
[0105] It should be noted that the greater the difference between the lateral vibration signals at different height levels, the worse the convective mixing uniformity of the solid powder.
[0106] Furthermore, by analyzing the lateral vibration difference in the i-th cycle and combining it with the distribution of peak values in the lateral vibration signals at all positions within the i-th cycle, the lateral vibration anomaly degree in the i-th cycle is obtained, specifically as follows:
[0107] Calculate the interquartile range of all maxima in each transverse vibration signal in time sequence; count the total number of maxima in each transverse vibration signal; calculate the sum of the products of the normalized interquartile range of the transverse vibration signals at all positions and the total number.
[0108] The normalized value of the product of the lateral vibration difference of the i-th period and the sum is taken as the lateral vibration anomaly degree of the i-th period.
[0109] In this embodiment, the Min-Max normalization method is used to obtain the normalized value of the interquartile range, and the Min-Max normalization method is used to obtain the normalized value of the product of the lateral vibration difference and the sum.
[0110] It should be noted that in the field of statistical analysis, count-scale joint statistics is a commonly used method for evaluating irregular peak distributions. To quantify the irregular multi-peak phenomenon caused by non-uniform convective mixing in transverse vibration signals, an evaluation method combining the number of maxima and the interquartile range is used. The greater the calculated transverse vibration anomaly, the worse the convective mixing uniformity of the solid powder.
[0111] Furthermore, during the homogeneous mixing of solid powders, when the uniformity of the shear mixing force experienced by the solid powders in the mixing chamber is poor, the local agglomeration or retention caused by the decrease in powder flowability will also trigger the elastic collision-recovery cycle of the powders in the axial direction, resulting in a severely irregular multi-peak distribution in the axial vibration signal; at the same time, due to the formation of the mixing accumulation zone or dead zone, the difference between the axial vibration signals at different height levels is stronger.
[0112] Calculate the axial vibration difference for the i-th period using the method for calculating lateral vibration difference. Calculate the axial vibration anomaly for the i-th period using the method for calculating lateral vibration anomaly.
[0113] In the process of homogenizing and mixing solid powders using a DCS homogenizing device with a built-in model predictive control algorithm, the higher the fluctuation amplitude of the torque voltage signal of the mixing chamber stirring shaft, the more severe the increase of the torque voltage signal, and the more severe the irregular multi-peak distribution phenomenon in the transverse vibration signal and axial vibration signal, the greater the difference in transverse vibration signal and axial vibration signal at different height levels in the mixing chamber, the more uneven the convective mixing force and shear mixing force the powder is subjected to in the mixing chamber, and the worse the homogenization and mixing state of the powder.
[0114] At this point, the softening coefficient in the model predictive control algorithm should be reduced to enable the control command to be issued quickly, thereby responding more quickly to the uneven local shear strength caused by the formation of dead zones or backflow zones in the mixing chamber. This allows for the use of greater shear energy to break up agglomerates and local accumulations inside the solid powder, improving the robustness of the solid powder homogenizing equipment to nonlinear disturbances and preventing the model predictive control algorithm from lingering in the low-energy region due to the softening trajectory, thus shortening the mixing time.
[0115] Based on the above analysis, a multi-attribute decision-making method is used to obtain the comprehensive anomaly coefficient for the i-th cycle, based on the torque anomaly, lateral vibration anomaly, and axial vibration anomaly. This comprehensive anomaly coefficient is used to evaluate the homogeneous mixing state of the solid powder within the i-th cycle. The larger the calculated comprehensive anomaly coefficient, the worse the homogeneous mixing state of the solid powder within the i-th cycle.
[0116] In this embodiment, the torque anomaly, lateral vibration anomaly, and axial vibration anomaly of the i-th cycle are input into the Technique for Order Preference by Similarity to Ideal Solution (TOPSIS), and the comprehensive score of the i-th cycle is output as the comprehensive anomaly coefficient of the i-th cycle. The weights of torque anomaly, lateral vibration anomaly, and axial vibration anomaly in the TOPSIS are obtained by the entropy weight method. Both the TOPSIS and entropy weight method are well-known technologies and will not be described in detail in this application. As another implementation method, based on the ability to jointly evaluate the homogeneous mixing state of the solid powder in the i-th cycle according to the torque anomaly, lateral vibration anomaly, and axial vibration anomaly, the implementer may adopt other existing feasible technologies, and this application does not impose any special restrictions.
[0117] It should be noted that in the field of statistical analysis, weighted fusion analysis is a commonly used method for comprehensive evaluation of multi-dimensional features. It involves extracting features from different data, assigning weights to each feature according to its importance, and then fusion of the weighted results of all features to obtain a comprehensive analysis result.
[0118] Calculate the comprehensive anomaly coefficient for each period according to the calculation method of the comprehensive anomaly coefficient for the i-th period.
[0119] Step 3: Adjust the softening coefficient in the model predictive control algorithm to control the mixing state of the solid powder.
[0120] Obtain the comprehensive anomaly coefficients of W homogenization devices in each historical period, and use the segmentation threshold of the comprehensive anomaly coefficients of W homogenization devices in all historical periods as the adjustment threshold for the softening coefficient.
[0121] In this embodiment, the value of W is 50. The value of W can be set by the implementer according to the actual situation, and this application does not impose any special restrictions.
[0122] In this embodiment, the Otsu threshold segmentation algorithm is used to obtain the segmentation threshold of the comprehensive anomaly coefficient. The Otsu threshold segmentation algorithm is a well-known technology and will not be described in detail in this application. As other implementation methods, based on the ability to obtain the segmentation threshold of the comprehensive anomaly coefficient, implementers may use other existing technologies, such as iterative threshold segmentation, global threshold segmentation, etc. This application does not impose any special restrictions.
[0123] During the homogenization and mixing of solid powder, if the comprehensive anomaly coefficient of each cycle is less than the softening coefficient adjustment threshold, the homogenization and mixing of the solid powder within each cycle is determined to be uniform, and the softening coefficient in the model predictive control algorithm is not adjusted. When the comprehensive anomaly coefficient of each cycle is greater than or equal to the softening coefficient adjustment threshold, the homogenization and mixing of the solid powder within each cycle is determined to be non-uniform, and the softening coefficient in the DCS built-in model predictive control algorithm of the homogenizing equipment needs to be adjusted. Specifically, the difference between 1 and the comprehensive anomaly coefficient of each cycle is recorded as the adjustment difference. The product of the preset initial value of the softening coefficient in the model predictive control algorithm and the adjustment difference is used as the softening coefficient in the model predictive control algorithm for the next cycle. A schematic diagram of the softening coefficient adjustment process is shown below. Figure 3 As shown.
[0124] All sampled values of the torque voltage signals acquired in each cycle are arranged in time sequence to form a torque voltage signal sequence for each cycle. Similarly, all sampled values of the lateral and axial vibration signals at each position in each cycle are arranged in time sequence to form lateral and axial vibration signal sequences for each position in each cycle. Using all torque voltage, lateral, and axial vibration signal sequences as input, a model predictive control algorithm (MMC) is employed to output a target power sequence for the stirring motor in the next cycle. The speed of the stirring motor is adjusted in real time via a frequency converter to ensure that its actual power keeps synchronized with the target power sequence. The softening coefficient in the MMC algorithm is the same as that used in the MMC algorithm for the next cycle. The target power sequence represents the power level that the stirring motor should achieve in the next cycle. The actual power of the stirring motor is adjusted in real time according to the target power sequence output by the MMC algorithm to ensure that the operating state of the stirring motor is consistent with the target state. This achieves precise control of the mixing process, optimizes the homogenization of solid powders, and improves the homogenization effect of solid powders.
[0125] In this embodiment, the initial value of the softening coefficient is 0.3, which is calculated from experimental data.
[0126] Step 4 shows the various parameters during the homogenization and mixing process of solid powder.
[0127] The DCS dashboard displays the torque voltage signal, lateral vibration signal, and axial vibration signal in real time for each cycle, and also displays the softening coefficient in the DCS built-in model predictive control algorithm of the homogenizing equipment in each cycle, helping relevant personnel to grasp the homogenization and mixing state of solid powder in a timely manner.
[0128] In summary, this application can quantify and analyze the characteristics of torque variation during the mixing process by calculating torque voltage fluctuation and low-frequency energy anomaly. Torque voltage fluctuation reflects the degree of change in the torque signal, while low-frequency energy anomaly reveals the variation law of the low-frequency component in the torque signal. By combining torque voltage fluctuation and low-frequency energy anomaly, the torque anomaly is obtained, which can more accurately characterize the degree of torque anomaly and reflect the mixing situation of solid powder in the mixing chamber from different perspectives.
[0129] Furthermore, by comprehensively considering the differences in vibration signals and peak distribution in each direction at different height positions, the convective and shear mixing states of solid powder in the mixing chamber can be fully evaluated. By integrating multi-dimensional vibration anomaly and torque anomaly, the homogeneous mixing state of solid powder can be evaluated more comprehensively and accurately, avoiding the one-sidedness of evaluation by a single index.
[0130] Furthermore, based on the homogenized mixing state of the solid powder obtained from the evaluation, the softening coefficient in the model predictive control algorithm is adjusted, thereby controlling the power of the stirring motor. This enables precise control of the mixing state of the solid powder, thereby improving the mixing effect. This dynamic adjustment mechanism based on the evaluation of the mixing state allows the homogenizing equipment to automatically adjust the control strategy according to the real-time changes in the mixing state, quickly respond to and optimize the mixing process, improve the robustness of the homogenizing equipment, and thus enhance the homogenization effect of the solid powder.
[0131] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to embodiments of this disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. In some alternative implementations, the functions marked in the blocks may occur in a different order than that shown in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. In the descriptions corresponding to the flowcharts and block diagrams in the accompanying drawings, the operations or steps corresponding to different blocks may also occur in a different order than disclosed in the description, and sometimes there is no specific order between different operations or steps. For example, two consecutive operations or steps may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. Each block in a block diagram and / or flowchart, and combinations of blocks in a block diagram and / or flowchart, can be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions.
[0132] It will be apparent to those skilled in the art that this application is not limited to the details of the exemplary embodiments described above, and that this application can be implemented in other specific forms without departing from its essential characteristics. Therefore, the embodiments described above should be considered exemplary and non-limiting in all respects.
Claims
1. A mixing control method for a homogenizing device for solid powder, characterized in that, The method includes the following steps: The torque voltage signal of the mixing shaft of the mixer in the solid powder homogenizing equipment is collected in real time, as well as the vibration signals of different directions at preset positions on the inner wall of the mixing chamber. The evaluation cycle of the homogeneous mixing state of the solid powder in the mixing chamber is preset; the torque voltage fluctuation of each cycle is obtained by the amplitude dispersion and rise of the torque voltage signal in each cycle, and the low frequency energy anomaly of each cycle is obtained by combining the periodic decline of the low frequency component of the torque voltage signal in each cycle and the energy ratio of the low frequency component, and the torque anomaly of each cycle is obtained. By measuring the differences in vibration signals in each direction at different heights within each cycle, the vibration differences in each direction within each cycle are obtained. Then, by combining the distribution of peak values in the vibration signals at all positions within each cycle in each direction, the vibration anomaly degree in each direction within each cycle is obtained. This is then fused with the torque anomaly degree to evaluate the homogeneous mixing state of the solid powder within each cycle. This allows for the adjustment of the softening coefficient in the model predictive control algorithm, thereby controlling the mixing state of the solid powder. The process for obtaining the torque voltage fluctuation is as follows: Calculate the dispersion of the amplitude of the torque voltage signal within each cycle; Each cycle is divided into windows. By analyzing the changes in the torque voltage signal between any two adjacent windows within each cycle, the degree of rise of the torque voltage signal within each cycle is measured. The torque voltage fluctuation is positively correlated with the results of the dispersion and the degree of rise, respectively. The process of obtaining the low-frequency energy anomaly is as follows: Obtain the low-frequency subband of the torque voltage signal within each window, and the periodic intensity of the energy values at all frequencies in the low-frequency subband; The difference in periodic intensity between each window and its adjacent next window within each period is denoted as the periodic difference. Calculate the cumulative energy value of the torque voltage signal at all frequencies in the low-frequency sub-band within each window; calculate the sum of the cumulative energy values of the torque voltage signal at all frequencies within each window; calculate the ratio of the cumulative value to the sum. The low-frequency energy anomaly is positively correlated with all the periodic differences in each period and also positively correlated with the sum of all the ratios in each period. The torque anomaly is the normalized result of the product of the torque voltage fluctuation and the low-frequency energy anomaly. The method for calculating the vibration difference is as follows: The positions are evenly distributed on the inner wall of the mixing chamber according to a pre-set row and column rule; The difference in vibration signal between any two adjacent rows and each column position in each direction is measured separately. The cumulative value of the measurement results of the difference in vibration signal between any two adjacent rows and each column position in each direction in each period is taken as the vibration difference in each direction in each period. The method for calculating the vibration anomaly degree is as follows: Calculate the interquartile range of all maxima in time for each vibration signal; count the total number of maxima for each vibration signal. For each direction, calculate the sum of the products of the normalized interquartile range of the vibration signals at all locations and the total number; The vibration anomaly degree is the normalized value of the product of the vibration difference and the sum; The process of evaluating the homogeneous mixing state of the solid powder in each period is as follows: Based on the torque anomaly and the vibration anomaly in all directions within each cycle, a multi-attribute decision method is used to obtain the comprehensive anomaly coefficient for each cycle, which is used to evaluate the homogeneous mixing state of the solid powder within each cycle. The method for evaluating the homogeneous mixing state of the solid powder in each period is as follows: Obtain the segmentation threshold of the comprehensive anomaly coefficient of a preset number of homogenized devices over multiple historical periods; If the overall anomaly coefficient of each period is less than the segmentation threshold, it is determined that the solid powder is homogeneously mixed in each period; otherwise, it is determined that the solid powder is not homogeneously mixed in each period. The method for adjusting the softening coefficient in the model predictive control algorithm is as follows: When the comprehensive anomaly coefficient of each period is greater than or equal to the segmentation threshold, the difference between 1 and the comprehensive anomaly coefficient of each period is recorded as the adjustment difference. The product of the preset initial value of the softening coefficient in the model predictive control algorithm and the adjustment difference is used as the softening coefficient in the model predictive control algorithm in the next period; otherwise, the softening coefficient in the model predictive control algorithm is not adjusted.
2. A homogenizing device for solid powder, employing the mixing control method of the homogenizing device for solid powder as described in claim 1, characterized in that, The device contains: The process monitoring module is used to collect the torque voltage signal of the mixing shaft of the mixer of the solid powder homogenizer in real time, as well as the vibration signals of different directions at preset positions on the inner wall of the mixing chamber. The data analysis module is used to preset the evaluation cycle of the homogeneous mixing state of solid powder in the mixing chamber; by measuring the amplitude dispersion and rise of the torque voltage signal in each cycle, the torque voltage fluctuation of each cycle is obtained; and by combining the periodic decline of the low-frequency component of the torque voltage signal in each cycle and the energy ratio of the low-frequency component, the low-frequency energy anomaly of each cycle is obtained, and the torque anomaly of each cycle is obtained. By measuring the differences in vibration signals in each direction at different heights within each cycle, the vibration differences in each direction within each cycle are obtained. The distribution of peak values in the vibration signals at all positions within each cycle in each direction is combined to obtain the vibration anomaly degree in each direction within each cycle. This is then fused with the torque anomaly degree to evaluate the homogeneous mixing state of the solid powder within each cycle. The DCS control module is used to adjust the softening coefficient in the model predictive control algorithm, thereby controlling the mixing state of solid powder.