Multi-stage quartz sand screening energy minimization system

Abstract

This invention belongs to the field of industrial process control technology. It discloses a system for minimizing energy consumption in multi-stage quartz sand screening. The system includes: a feature identification module for acquiring the nonlinear hysteresis feature vectors of each stage of the vibrating screen during the low-load window, dividing the amplitude ranges of high-amplitude stable branches and low-amplitude stable branches, and forming a process switching reference table; an entropy production mechanism establishment module for constructing a multi-scale entropy production rate generation mechanism for each stage of the vibrating screen; a branch transition module for triggering nonlinear resonance transitions when it is determined that the current operating point and the target operating point do not belong to the same stable branch; an optimal path planning module for planning the optimal reference path and forming a parameter sequence evolving towards the target operating point; and a closed-loop correction module for updating the multi-scale entropy production rate generation mechanism, performing closed-loop tracking of the optimal reference path, and achieving optimal control of global energy consumption during process switching.

Description

Technical Field

[0001] This invention relates to the field of industrial process control technology, and more specifically, to a system for minimizing energy consumption in multi-stage quartz sand screening. Background Technology

[0002] Quartz sand is a core raw material in glass manufacturing, building materials processing and other fields. Its particle size classification accuracy and processing energy consumption directly affect the quality of downstream products. Multi-stage screening can achieve efficient classification of quartz sand of different particle sizes by connecting various vibrating screens in series, and has been widely used in the large-scale production and processing of quartz sand.

[0003] Existing multi-stage quartz sand screening control technologies generally employ linear control strategies. This involves combining constant-frequency drive with fixed-parameter formulations using frequency converters to call preset combinations of excitation frequency, amplitude, and other parameters based on the target finished product specifications, maintaining the screening production line's ability to meet the particle size classification requirements of routine production tasks. However, existing multi-stage screening control strategies typically simplify the dynamic response of vibrating screens with material loads to a linear characteristic, setting the excitation force and vibrating screen amplitude to change proportionally, without considering the nonlinear effects of material loads during actual screening. In actual operating conditions, vibrating screens with quartz sand material loads exhibit significant nonlinear stiffness characteristics, and their amplitude-frequency characteristic curves show typical hysteresis near the resonance region. This means that there are two possible stable amplitude branches within a specific frequency range, and existing linear control strategies cannot identify this nonlinear hysteresis characteristic. Consequently, when switching production tasks, the frequency can only be passively adjusted using linear interpolation, easily getting stuck in low-efficiency branches or crossing unstable regions, causing violent oscillations. This makes it difficult to achieve optimal energy consumption control during process switching and steady-state operation, hindering further improvement in the overall energy efficiency of multi-stage screening production lines.

[0004] In view of this, the present invention proposes a system for minimizing energy consumption in multi-stage screening of quartz sand to solve the above problems. Summary of the Invention

[0005] To overcome the aforementioned deficiencies of the prior art and to achieve the above objectives, the present invention provides the following technical solution: a multi-stage quartz sand screening energy-saving system, comprising:

[0006] The feature identification module is used to obtain the nonlinear hysteresis feature vectors of each level of vibrating screen during the low load window period, divide the amplitude range of the high amplitude stable branch and the low amplitude stable branch, and form a process switching reference table.

[0007] The entropy production mechanism establishment module is used to construct a multi-scale entropy production rate generation mechanism for vibrating screens at all levels within a four-dimensional control parameter space.

[0008] The branch transition module is used to parse the current operating point when the process switching command is executed, extract the target operating point and its stable branch corresponding to the current operating point from the process switching reference table, and trigger a nonlinear resonance transition if it is determined that the current operating point and the target operating point do not belong to the same stable branch.

[0009] The optimal path planning module is used to call the multi-scale entropy yield generation mechanism when the resonance transition is completed or the same stable branch is determined. It plans the optimal reference path in the four-dimensional control parameter space that minimizes the total entropy yield of the multi-stage sieving along the path integral and forms a parameter sequence that evolves towards the target working point.

[0010] The closed-loop correction module is used to analyze the deviation between the actual total entropy yield and the theoretical total entropy yield during the evolution process, update the multi-scale entropy yield generation mechanism, and use the updated multi-scale entropy yield generation mechanism as the optimization target to perform closed-loop tracking of the optimal reference path.

[0011] Furthermore, the nonlinear hysteresis characteristic vectors of each stage of the vibrating screen during the low-load window are obtained, including:

[0012] An upward sweep frequency excitation signal with linearly increasing frequency and a downward sweep frequency excitation signal with linearly decreasing frequency are applied to each level of vibrating screen during the low load window period. The vibration acceleration signals of each level of vibrating screen during the sweep frequency are collected synchronously and converted into time and frequency to obtain the upward amplitude-frequency characteristic curve and the downward amplitude-frequency characteristic curve of each level of vibrating screen.

[0013] The peak amplitude and corresponding frequency are extracted from the uplink amplitude-frequency characteristic curve and used as the maximum amplitude and resonant frequency of the uplink path, respectively; the amplitude value and frequency before the amplitude drops sharply in the downlink amplitude-frequency characteristic curve are identified and used as the maximum amplitude and resonant frequency of the downlink path, respectively.

[0014] The maximum amplitude of the upward path, the resonant frequency of the upward path, the maximum amplitude of the downward path, and the resonant frequency of the downward path are combined to form the nonlinear hysteresis characteristic vector of each level of the vibrating screen.

[0015] Furthermore, a process switchover reference table is created, including:

[0016] The bistable frequency range is defined by the resonant frequencies of the uplink and downlink paths;

[0017] Within the bistable frequency range, the maximum amplitudes of the downlink and uplink paths are used as the lower and upper bounds for determining the amplitude range of the high-amplitude stable branch, respectively. The amplitude value at which the amplitude suddenly drops and then re-stabilizes in the downlink amplitude-frequency characteristic curve is extracted and used, along with the minimum allowable stable amplitude, as the upper and lower bounds for determining the amplitude range of the low-amplitude stable branch, respectively.

[0018] A mapping relationship database is established based on historical quartz sand screening data; the mapping relationship database records the mapping relationship between the finished quartz sand specifications and the corresponding four-dimensional control parameters;

[0019] The amplitude of each finished product specification is compared with the amplitude range of each stable branch. The stable branch obtained from the comparison is associated with the corresponding four-dimensional control parameters to form a process switching reference table.

[0020] Furthermore, a multi-scale entropy yield generation mechanism is constructed for vibrating screens at all levels, including:

[0021] Within a four-dimensional control parameter space consisting of frequency, amplitude, vibration direction angle and feeding speed, a multi-scale entropy yield generation mechanism is constructed for each level of vibrating screen, including particle-scale entropy yield sub-mechanisms and screen-scale entropy yield sub-mechanisms.

[0022] The particle-scale entropy yield sub-mechanism is used to obtain the particle-scale entropy yield, which characterizes the inelastic collision and friction energy dissipation of quartz sand particles, by logical deduction based on the measured collision recovery coefficient and rolling friction coefficient of quartz sand particles, combined with the vibration direction angle and feeding speed.

[0023] The screen-scale entropy yield sub-mechanism is used to analyze the loss factor of screen material measured by vibrating screens at each level, combined with the metal structural stiffness, vibration frequency and amplitude of each level of vibrating screen, to obtain the screen-scale entropy yield that characterizes the internal friction and elastic hysteresis dissipation of screen material at each level of vibrating screen.

[0024] Furthermore, extract the target working point and its stable branch corresponding to the current working point, including:

[0025] The received process switching command is parsed to obtain the target finished product specifications; the vibration frequency, amplitude, vibration direction angle and feeding speed of each level of vibrating screen are taken as the current working point; based on the vibration frequency and amplitude of the current working point, combined with the nonlinear hysteresis characteristic vector, the stable branch to which the current working point belongs is determined.

[0026] Retrieve the stable branch corresponding to the target finished product specification from the process switching reference table, as well as the target four-dimensional control parameters that serve as the corresponding target working point; determine whether the stable branch to which the current working point and the target working point belong are the same. If they are not the same, trigger a nonlinear resonant transition.

[0027] Furthermore, triggering nonlinear resonant transitions includes:

[0028] Calculate the average value of the resonance frequencies of the upward and downward paths of each level of vibrating screen as the crossing frequency; adjust the excitation frequency from the vibration frequency of the current working point to the crossing frequency according to the first preset rate of change, and apply a transient excitation force pulse to the excitation motor synchronously when it is determined that the excitation frequency crosses the crossing frequency.

[0029] After the transient excitation force pulse is applied, the response amplitude of each level of the vibrating screen is monitored; when each response amplitude jumps to the amplitude range of the stable branch to which the target working point belongs, the excitation frequency is adjusted to the vibration frequency corresponding to the target working point with the second preset change rate, so as to complete the resonance transition and frequency alignment from the current working point to the stable branch to which the target working point belongs.

[0030] Furthermore, an optimal reference path is planned that minimizes the integral of the total entropy yield of the multi-stage screening process along the path, including:

[0031] Establish inter-stage material particle size distribution transmission constraints; under the constraint of inter-stage material particle size distribution transmission, the multi-scale entropy yield obtained by the multi-scale entropy yield generation mechanism is the cost target. The current working point is taken as the planning start point and the target working point is taken as the planning end point. Search and plan in the four-dimensional control parameter space to obtain the optimal reference path that minimizes the total entropy yield of multi-stage screening along the path.

[0032] The optimal reference path is discretized into a parameter sequence, which includes the vibration frequency sequence, amplitude sequence, vibration direction angle sequence, and feed rate sequence from the current working point to the target working point.

[0033] Furthermore, the deviation between the actual total entropy yield and the theoretical total entropy yield during the multi-stage sieving process is analyzed, including:

[0034] The real-time input power of the drive motors of each level of vibrating screen and the real-time finished product output fed back by each belt scale are obtained and correlated to obtain the actual total entropy yield of multi-stage screening.

[0035] The theoretical total entropy yield obtained from the analysis scale entropy yield generation mechanism is compared with the actual total entropy yield from multi-stage sieving, and the multi-scale entropy yield generation mechanism is updated based on the deviation results.

[0036] Furthermore, the multi-scale entropy yield generation mechanism is updated based on the bias results, including:

[0037] The absolute value of the deviation between the theoretical total entropy yield and the actual total entropy yield of multi-stage screening is analyzed and compared with the preset deviation judgment threshold. If the absolute value of the deviation is greater than the deviation judgment threshold, it is determined that there is a deviation. The collision recovery coefficient and rolling friction coefficient of quartz sand particles are corrected according to the direction and amplitude of the deviation.

[0038] If the absolute value of the deviation is less than or equal to the offset determination threshold, it is determined that there is no offset, and the multi-scale entropy yield generation mechanism remains unchanged.

[0039] Furthermore, closed-loop tracking of the optimal reference path is performed, including:

[0040] Obtain the target output of the current production task; with the updated multi-scale entropy yield generation mechanism as the optimization objective, in each rolling optimization time domain, solve for the optimal parameter sequence that minimizes the cumulative value of the total entropy yield of multi-stage screening in the prediction time domain and satisfies that the finished product output is not lower than the target output.

[0041] The first step control increment in the optimal parameter sequence is superimposed on the vibration frequency, amplitude, vibration direction angle and feeding speed of the current level of vibrating screen and continuously evolved; the current working point is updated according to the real-time response of each level of vibrating screen, and closed-loop rolling optimization control is executed until the closed-loop tracking of the optimal reference path is completed.

[0042] The technical effects and advantages of the present invention's energy-saving system for multi-stage quartz sand screening are as follows:

[0043] 1. This invention applies bidirectional frequency sweep excitation to each level of vibrating screen during a low-load window period, extracts the resonant frequency and maximum amplitude of the upward and downward paths to form a nonlinear hysteresis feature vector, delineates the bistable frequency range and defines the amplitude judgment boundary of the high and low amplitude stable branches, and stores the process parameters of each finished product specification and its corresponding stable branch as a process switching reference table. This realizes an explicit quantitative expression of the nonlinear hysteresis characteristics in the quartz sand screening process, enabling accurate identification of the branch to which any working point belongs during the screening process. This fundamentally avoids the branch misjudgment caused by the traditional linear strategy ignoring hysteresis characteristics, and provides a reliable dynamic perception basis for subsequent transition decisions.

[0044] 2. This invention analyzes process switching commands and achieves optimal global energy consumption control during process switching through phased collaboration of branch transition control and minimum entropy production path planning. When it is determined that the current operating point and the target operating point belong to different stable branches, a nonlinear resonant transition is triggered. The system is driven to cross to the target branch with low energy consumption by combining the crossing frequency with transient excitation force pulses. This solves the problem of traditional control passively staying in low-energy-efficiency branches and continuously high energy consumption. After the transition is completed, the optimal reference path that minimizes the total entropy production rate of multi-stage screening along the path is planned in the four-dimensional control parameter space. This expands the energy consumption optimization dimension from the steady-state operating point to the entire dynamic transition process. The transition-cruise collaborative architecture formed by the organic integration of the two achieves optimal global energy consumption control during large-scale process switching, significantly improving the operating energy efficiency of multi-stage quartz sand screening and the process adaptability to different finished product specifications. The closed-loop correction module monitors and corrects the multi-scale entropy production rate generation mechanism in real time, ensuring that the actual operating trajectory always closely follows the optimal reference path, further consolidating the energy-saving effect and operational stability of the process switching process. Attached Figure Description

[0045] Figure 1 This is a system principle diagram of the quartz sand multi-stage screening energy-saving system of the present invention;

[0046] Figure 2 This is a flowchart illustrating the process of determining whether a nonlinear resonant transition is triggered according to the present invention.

[0047] Figure 3 This is a flowchart illustrating the nonlinear resonant transition triggering of the present invention. Detailed Implementation

[0048] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0049] Example 1, please refer to Figure 1 , Figure 2 and Figure 3 As shown, the quartz sand multi-stage screening energy-saving system described in this embodiment includes:

[0050] The feature identification module is used to obtain the nonlinear hysteresis feature vectors of each level of vibrating screen during the low load window period, divide the amplitude range of the high amplitude stable branch and the low amplitude stable branch, and form a process switching reference table.

[0051] Obtain the nonlinear hysteresis characteristic vectors of each level of vibrating screen during the low load window period, including:

[0052] An upward sweep excitation signal with linearly increasing frequency and a downward sweep excitation signal with linearly decreasing frequency are applied to each stage of the vibrating screen during the low-load window period. The low-load window period refers to a specific time period during the normal operation cycle of the multi-stage quartz sand screening production line when the instantaneous quartz sand feed rate of the feed belt is less than 20% of the rated quartz sand feed rate or when there is no material passing through. By performing sweep frequency identification during the low-load window period, the signal-to-noise ratio of the nonlinear characteristics of each stage of the vibrating screen can be maximized.

[0053] The purpose of applying the frequency sweep excitation signal is to drive the excitation motor to produce forced vibration of each stage of the vibrating screen. The frequency sweep excitation signal and the frequency sweep excitation signal have their frequency scanning ranges set according to the factory design resonant frequency parameters of each stage of the vibrating screen; the frequency scanning range can completely cover the range of resonant frequency shift towards higher frequencies caused by stiffness hardening nonlinear effects under the load condition of quartz sand material.

[0054] For example, the frequency of the uplink sweep excitation signal increases linearly from 45Hz to 110Hz, with a sweep rate of 1Hz / s. The frequency of the downlink sweep excitation signal decreases linearly from 110Hz to 45Hz, with a sweep rate of 1Hz / s.

[0055] The vibration acceleration signals of each level of the vibrating screen are synchronously acquired and converted into time-frequency signals to obtain the upward amplitude-frequency characteristic curves and downward amplitude-frequency characteristic curves of each level of the vibrating screen.

[0056] Vibration acceleration signals are synchronously acquired by acceleration sensors installed on the screen frames of each level of the vibrating screen.

[0057] The uplink amplitude-frequency response curve characterizes the process during uplink frequency sweep where the amplitude jumps from a low amplitude branch to a high amplitude branch via a sudden increase; the downlink amplitude-frequency response curve characterizes the process during downlink frequency sweep where the amplitude jumps from a high amplitude branch to a low amplitude branch via a sudden decrease.

[0058] The specific process of time-frequency conversion is as follows: Using Fourier transform, the vibration acceleration signals corresponding to the upward sweep excitation signals of each stage of the vibrating screen are converted from the time domain to the frequency domain, obtaining the response amplitude corresponding to each frequency point. The response amplitude values ​​corresponding to each frequency point during the upward sweep are connected in ascending order of frequency to form the upward amplitude-frequency characteristic curve. The vibration acceleration signals corresponding to the downward sweep excitation signals are sampled and processed in the same way, and the response amplitude values ​​corresponding to each frequency point during the downward sweep are connected in descending order of frequency to form the downward amplitude-frequency characteristic curve.

[0059] The peak amplitude and corresponding frequency are extracted from the upward amplitude-frequency characteristic curve and used as the maximum amplitude and resonant frequency of the upward path, respectively. The resonant frequency of the upward path characterizes the critical frequency at which the vibrating screen jumps from the low-amplitude stable branch to the high-amplitude stable branch. The maximum amplitude of the upward path characterizes the maximum amplitude that each stage of the vibrating screen can achieve on the high-amplitude stable branch.

[0060] Identify the amplitude value and frequency before the sudden drop in amplitude in the downlink amplitude-frequency characteristic curve, and use them as the maximum amplitude and resonant frequency of the downlink path, respectively.

[0061] Arrange the response amplitudes in the downlink amplitude-frequency response curve into a sequence in descending order of frequency. Starting from the second data point in the sequence, calculate the rate of change of amplitude between the current data point and the previous data point. When the absolute value of the rate of change of amplitude at a certain data point first exceeds the sum of the average of the absolute values ​​of all amplitude rates of change and a certain number of standard deviations (e.g., 3 times), mark that data point as the amplitude abrupt threshold frequency point. The data point preceding the amplitude abrupt threshold frequency point and its corresponding frequency are taken as the maximum amplitude and resonant frequency of the downlink path, respectively.

[0062] The resonant frequency of the downlink path characterizes the critical frequency at which a transition occurs from a high-amplitude stable branch to a low-amplitude stable branch. The resonant frequency of the uplink path must be greater than that of the downlink path.

[0063] The maximum amplitude of the upward path, the resonant frequency of the upward path, the maximum amplitude of the downward path, and the resonant frequency of the downward path are combined to form the nonlinear hysteresis characteristic vector of each stage of the vibrating screen. For each stage of the vibrating screen, the maximum amplitude of the upward path, the resonant frequency of the upward path, the maximum amplitude of the downward path, and the resonant frequency of the downward path are packaged into a four-element array structure in sequence and bound to the corresponding stage identifier of the vibrating screen to form the corresponding nonlinear hysteresis characteristic vector.

[0064] Nonlinear hysteresis eigenvectors are used to transform the nonlinear hysteresis characteristics of vibrating screens, which are ignored in traditional control strategies, into identifiable, quantifiable, and callable structured data.

[0065] A process switchover reference table is generated, including:

[0066] The bistable frequency range is defined by using the resonant frequencies of the upward and downward paths as boundaries. The resonant frequency of the downward path is taken as the lower limit of the bistable frequency range, and the resonant frequency of the upward path is taken as the upper limit of the bistable frequency range, thus obtaining the bistable frequency range of each stage of the vibrating screen.

[0067] The bistable frequency range is used to characterize the resonant hysteresis region exhibited in the frequency response curve of each level of vibrating screen under the action of quartz sand load and affected by nonlinear dynamic characteristics. Within the resonant hysteresis region, each level of vibrating screen exhibits two incompatible steady motion states for the same excitation frequency: a high-amplitude stable branch and a low-amplitude stable branch.

[0068] For example, the resonant frequency of the upward path of a primary vibrating screen is 92.5 Hz, and the resonant frequency of the downward path is 87.0 Hz. Therefore, the bistable frequency range of this primary vibrating screen is [87.0, 92.5].

[0069] Within the bistable frequency range, the maximum amplitudes of the downward and upward paths are used as the lower and upper bounds for determining the amplitude range of the high-amplitude stable branch, respectively. In other words, the closed interval between the lower and upper bounds is defined as the amplitude range of the high-amplitude stable branch of the vibrating screen at this stage.

[0070] The high-amplitude stable branch indicates that the vibrating screen operates within a bistable frequency range, where the screen frame vibration amplitude is at a relatively high level and in a stable working state. Within the high-amplitude stable branch, the quartz sand on the screen surface can achieve a large throwing acceleration and throwing height, resulting in a high degree of particle looseness, significant stratification, a high probability of fine particles passing through the screen, and a high level of screening efficiency.

[0071] Extract the amplitude value at which the amplitude suddenly drops and then stabilizes again from the downlink amplitude-frequency characteristic curve, and use it, along with the minimum allowable stable amplitude, as the upper and lower bounds for determining the amplitude range of the low-amplitude stable branch, respectively.

[0072] "Re-stabilization" refers to the period during which the vibrating screen transitions from a high-amplitude vibration state to a low-amplitude vibration state, during which the transient irregular fluctuations caused by energy dissipation and stiffness switching cease, and the vibrating screen enters a quasi-static steady motion stage.

[0073] Starting from the frequency point where the amplitude suddenly drops to its critical threshold, a sliding analysis window is constructed along the direction of frequency decrease. If the standard deviation of the amplitude values ​​in the next sliding window is less than or equal to half of the standard deviation of the amplitude values ​​in the previous sliding window, the amplitude is determined to have re-entered a steady state. The arithmetic mean of all amplitude values ​​in the next sliding analysis window is taken as the amplitude value at which the amplitude stabilizes again after the sudden drop.

[0074] The minimum allowable stable amplitude refers to the minimum amplitude value at which the screen frame can maintain identifiable stable vibration, distinct from a completely static state, under normal operating conditions. The minimum allowable stable amplitude is taken as the arithmetic mean of several consecutive amplitude values ​​at the lowest frequency end of the downward amplitude-frequency characteristic curve.

[0075] The low-amplitude stable branch refers to a stable operating state where the screen frame vibration amplitude is at a low level when the vibrating screen is running within the same excitation frequency range. Under the low-amplitude stable branch, the quartz sand on the screen surface receives insufficient throwing acceleration, the particle group is not sufficiently loose, the stratification effect is poor, and fine particles have difficulty effectively passing through the material layer to reach the screen surface, resulting in low screening efficiency. To maintain the same finished product output, it is often necessary to increase the feeding time or increase the feeding speed, leading to a significant increase in energy consumption per unit output.

[0076] A mapping database is established based on historical quartz sand screening data. This database records the mapping relationship between finished quartz sand specifications and corresponding four-dimensional control parameters (i.e., vibration frequency, amplitude, vibration direction angle, and feed rate). Historical quartz sand screening data refers to data recorded for different finished quartz sand specifications during stable operation periods of the multi-stage quartz sand screening production line. Each historical quartz sand screening data entry includes the finished quartz sand specification identifier, vibration frequency, amplitude, vibration direction angle, feed rate, and energy consumption per unit output.

[0077] It should be explained that the values ​​of the four-dimensional control parameters for each finished product specification in the mapping relationship library are the values ​​of the four-dimensional control parameters corresponding to the lowest energy consumption per unit output record in the historical quartz sand screening data.

[0078] For example, the mapping database stores:

[0079] Quartz sand finished product specifications (20-40 mesh): vibration frequency (85.0Hz), amplitude (0.18mm), vibration direction angle (40°), feeding speed (22.5t / h);

[0080] Quartz sand finished product specifications (40-70 mesh): vibration frequency (92.0Hz), amplitude (0.12mm), vibration direction angle (35°), feeding speed (18.0t / h).

[0081] The amplitude of each finished product specification is compared with the amplitude range of each stable branch. The stable branch obtained from the comparison is associated with the corresponding four-dimensional control parameters to form a process switching reference table.

[0082] The entropy production mechanism establishment module is used to construct a multi-scale entropy production rate generation mechanism for vibrating screens at all levels within a four-dimensional control parameter space.

[0083] To construct a multi-scale entropy yield generation mechanism for vibrating screens at all levels, including:

[0084] Within a four-dimensional control parameter space consisting of frequency, amplitude, vibration direction angle, and feeding speed, a multi-scale entropy yield generation mechanism is constructed for each level of vibrating screen, including particle-scale entropy yield sub-mechanisms and screen-scale entropy yield sub-mechanisms.

[0085] The particle-scale entropy yield sub-mechanism is used to deduce the particle-scale entropy yield, which characterizes the energy dissipation from inelastic collisions and friction of quartz sand particles, by logically deducing the measured collision recovery coefficient and rolling friction coefficient of quartz sand particles in combination with the vibration direction angle and feed rate. The higher the particle-scale entropy yield, the greater the ineffective energy consumption used to overcome the disordered motion between particles.

[0086] Currently, the vibration direction angle and feeding speed of each level of vibrating screen are obtained in real time through corresponding sensors. The vibration direction angle is the angle between the vibration direction and the horizontal direction; the feeding speed is the instantaneous speed at which the quartz sand enters the vibrating screen.

[0087] Read the collision recovery coefficient and rolling friction coefficient of quartz sand particles that have been calibrated and stored offline. Obtain the average mass of a single quartz sand particle within a specified particle size range from the target mining area offline.

[0088] The logic for obtaining the particle-scale entropy yield is as follows:

[0089] By calculating the physical relationships between amplitude, vibration frequency, and vibration direction angle, the velocity components of the screen surface motion along and perpendicular to the screen surface are obtained. Combined with the feeding speed, the average relative velocity between particles is calculated. The average collision frequency of quartz sand per unit time is calculated from the feeding speed, the average mass of quartz sand particles, and the amount of quartz sand particle clusters remaining on the screen surface.

[0090] The kinetic energy loss ratio of a single inelastic collision is obtained based on the collision recovery coefficient of quartz sand particles (kinetic energy loss ratio = 1 - the square of the collision recovery coefficient of quartz sand particles); the total kinetic energy loss caused by inelastic collisions per unit time is calculated, and the total kinetic energy loss = average mass × the square of the average relative velocity between particles × the average collision frequency per unit time × the kinetic energy loss ratio (both are dimensionless).

[0091] Based on the rolling friction coefficient of quartz sand particles, the average normal pressure between particles (i.e., the average pressure perpendicular to the contact surface generated on the contact surface between quartz sand particles deep within the material layer due to the gravity of the material column above, obtained by multiplying the measured average thickness of the quartz sand layer by the bulk density of the quartz sand), and the vibration velocity component of the quartz sand particles on the screen surface, the total frictional energy loss per unit time due to rolling and sliding friction is calculated. The total kinetic energy loss per unit time is algebraically added to the total frictional energy loss to obtain the total dissipated power of inelastic collisions and friction between quartz sand particles per unit time. The quotient of the total dissipated power divided by the measured absolute ambient temperature is taken as the particle-scale entropy yield.

[0092] The screen-scale entropy yield sub-mechanism is used to analyze the loss factor of screen material measured by vibrating screens at each level, combined with the metal structural stiffness, vibration frequency and amplitude of each level of vibrating screen, to obtain the screen-scale entropy yield that characterizes the internal friction and elastic hysteresis dissipation of screen material at each level of vibrating screen.

[0093] The screen material loss factor is a dimensionless parameter representing the proportion of elastic strain energy converted into heat energy by the screen surface material of a vibrating screen within one vibration cycle. It is used to calculate the energy dissipated per unit volume of screen material per unit time due to internal friction and elastic hysteresis. The screen material loss factor is obtained by measuring samples of screen material from various levels of vibrating screens offline using a dynamic mechanical analyzer.

[0094] Metal structural stiffness represents the ability of the screen frame of each level of vibrating screen to resist elastic deformation when subjected to external forces. The greater the metal structural stiffness, the greater the base amount of elastic strain energy stored inside the material under the same amplitude, which is the source of energy dissipation.

[0095] The expression for obtaining the entropy yield at the sieve scale, derived through the sieve scale entropy yield sub-mechanism, is as follows:

[0096]

[0097] in, This is expressed as the entropy yield of the screen size of the s-th level vibrating screen (s=1,2...).

[0098] This represents the effective vibration volume of the s-th level vibrating screen mesh. The effective vibration volume is obtained by continuously multiplying the length, width, and wire diameter of the vibrating screen mesh. The length, width, and wire diameter of the vibrating screen mesh are all factory design parameters and are pre-stored in the memory.

[0099] The loss factor of the screen material was determined. This represents the metal structural stiffness of the s-th level vibrating screen; the metal structural stiffness is obtained through a static loading test on the screen frame of the vibrating screen. This is expressed as the amplitude of the s-th level vibrating screen; This is expressed as the vibration frequency of the s-th level vibrating screen; It represents the absolute ambient temperature, which is obtained in real time by temperature sensors installed on the screening production line.

[0100] In the expression for obtaining the entropy yield at the screen scale, all parameters participate in the calculation as dimensionless relative values ​​that have been normalized.

[0101] The branch transition module is used to parse the current operating point when the process switching command is executed. It extracts the target operating point and its stable branch corresponding to the current operating point from the process switching reference table. If it is determined that the current operating point and the target operating point do not belong to the same stable branch, a nonlinear resonance transition is triggered.

[0102] Extract the target working point and its stable branch corresponding to the current working point, including:

[0103] The received process switching command is parsed to obtain the target finished product specification. The process switching command is a control command that instructs the current production task to switch from one type of quartz sand finished product specification to another. The process switching command includes a command type identifier, a target finished product specification identifier, and a command issuance timestamp.

[0104] For example, after completing the process switch to producing 20-40 mesh quartz sand in a batch according to the production schedule, the multi-stage quartz sand screening production line automatically sends a process switch command to the controller. The command content is as follows: the command type is identified as "process switch", the target finished product specification is identified as "40-70 mesh", and the command issuance timestamp is "February 21, 2026, 14:30:00".

[0105] The vibration frequency, amplitude, vibration direction angle, and feeding speed of each level of vibrating screen are taken as the current working point; based on the vibration frequency and amplitude of the current working point, combined with the nonlinear hysteresis characteristic vector, the stable branch to which the current working point belongs is determined.

[0106] It should be explained that if the vibration frequency of the current operating point is outside the bistable frequency range, the branch comparison and resonance transition control are skipped, and the optimal reference path planning is performed directly. If the vibration frequency of the current operating point is within the bistable frequency range, but the current amplitude does not meet the requirements of the two types of stable branches, the current operating point is determined to be in a transitional state, and the system waits for the vibration frequency, amplitude, vibration direction angle, and feeding speed collected in the next sampling cycle to make a judgment.

[0107] Retrieve the stable branch corresponding to the target finished product specification from the process switching reference table, as well as the target four-dimensional control parameters that serve as the corresponding target working point.

[0108] Determine whether the current operating point and the target operating point belong to the same stable branch. If they are not the same, trigger a nonlinear resonant transition.

[0109] If the current working point and the target working point belong to different stable branches, it indicates that the vibrating screens at each level are in an operating branch that does not match the energy efficiency required by the target finished product specifications.

[0110] For example, the target finished product specification corresponds to a high amplitude stable branch to obtain better energy efficiency, but the current working point is stuck in a low amplitude stable branch; or the target finished product specification corresponds to a low amplitude stable branch to meet specific particle size classification accuracy or avoid excessive screen wear, but the current working point is stuck in a high amplitude stable branch, which is a working state that does not match the process requirements.

[0111] Nonlinear resonant transition refers to the process where, when it is determined that the current operating point and the target operating point belong to different stable branches, the vibrating screen needs to be driven to jump from the current stable branch to the target stable branch by adjusting the excitation frequency and applying a brief additional excitation force. Nonlinear resonant transition enables active matching of energy efficiency states with minimal energy cost, while avoiding severe amplitude oscillations and mechanical shocks caused by uncontrollable passive jumps due to external random disturbances.

[0112] Triggering nonlinear resonant transitions includes:

[0113] Calculate the average value of the resonant frequencies of the upward and downward paths of each level of vibrating screen, and use it as the crossing frequency.

[0114] The frequency span is located at the center of the bistable frequency range. At this frequency, the amplitude difference between the high-amplitude stable branch and the low-amplitude stable branch reaches its maximum, and the frequency is in the most unstable critical state. At this time, the energy cost required to apply external transient excitation is the minimum.

[0115] The excitation frequency is adjusted from the vibration frequency of the current working point to the crossing frequency according to the first preset rate of change. When it is determined that the excitation frequency crosses the crossing frequency (that is, when the actual value of the excitation frequency is detected to enter the crossing frequency for the first time), a transient excitation force pulse is synchronously applied to the excitation motor.

[0116] The first preset rate of change is usually a small value, typically ranging from 0.5 Hz / s to 1 Hz / s, so that the vibration frequency adjustment process is approximately a quasi-static process. This ensures that when the vibration frequency approaches the crossover frequency, the vibrating screen can maintain operation on the current stable branch, avoiding premature triggering of uncontrollable passive jumps due to dynamic effects.

[0117] It should be explained that the transient excitation force pulse is an excitation voltage signal applied by the frequency converter to the excitation motor with a very short duration and an amplitude higher than the current operating level (usually set to 1.2 to 1.5 times the amplitude of the current operating point). It is used to provide a brief external energy impact to the vibrating screen to help the vibrating screen overcome the dynamic barrier of the unstable branch in the middle of the hysteresis loop.

[0118] After the transient excitation force pulse is applied, the response amplitude of each level of the vibrating screen is monitored; when each response amplitude jumps to the amplitude range of the stable branch to which the target working point belongs, the excitation frequency is adjusted to the vibration frequency corresponding to the target working point with the second preset change rate, so as to complete the resonance transition and frequency alignment from the current working point to the stable branch to which the target working point belongs.

[0119] It should be explained that the second preset rate of change is used in the frequency alignment stage after the successful transition. In this stage, each level of the vibrating screen is already within the target stable branch range, and further adjustment of the vibration frequency is only carried out within the same stable branch, without involving branch crossing. Therefore, the value of the second preset rate of change can be greater than the first preset rate of change (e.g., the range is usually set to 1Hz / s to 2Hz / s) to shorten the overall transition time of the process switching.

[0120] The optimal path planning module is used to call the multi-scale entropy yield generation mechanism when a resonant transition is completed or when it is determined to belong to the same stable branch. It plans the optimal reference path in the four-dimensional control parameter space that minimizes the total entropy yield of the multi-stage sieving along the path integral and forms a parameter sequence that evolves toward the target working point.

[0121] The optimal reference path is planned to minimize the integral of the total entropy yield of the multi-stage screening process along the path, including:

[0122] Establish inter-stage material particle size distribution transfer constraints. These constraints, based on the process characteristics of multi-stage cascaded screening, limit the feed particle size distribution of each subsequent vibrating screen to be entirely determined by the oversize particle size distribution of the preceding vibrating screen. Furthermore, the qualified undersize material from each vibrating screen directly enters the finished product collection stage and does not participate in subsequent screening stages. When planning the optimal reference path within the four-dimensional control parameter space, all parameter point sequences must satisfy this inter-stage transfer relationship to ensure that the planned optimal reference path conforms to the actual screening process logic and is engineering-executable.

[0123] Under the constraint of material particle size distribution transmission between stages, the multi-scale entropy yield obtained by the multi-scale entropy yield generation mechanism is the cost target. The current working point is taken as the planning start point and the target working point is taken as the planning end point. The search planning is carried out in the four-dimensional control parameter space to obtain the optimal reference path that minimizes the total entropy yield of multi-stage screening along the path.

[0124] The specific method for search planning is as follows: The continuous trajectory from the planning start point to the planning end point in the four-dimensional control parameter space is discretized into a sequence of several parameter points. Each parameter point contains the vibration frequency, amplitude, vibration direction angle, and feed rate of each level of vibrating screen at the current moment. The particle-scale entropy yield and screen-scale entropy yield of each level of vibrating screen at that parameter point are calculated separately using a multi-scale entropy yield generation mechanism, and then algebraically added to obtain the multi-scale entropy yield of the corresponding level of vibrating screen. The multi-scale entropy yields of each level of vibrating screen are then superimposed to obtain the total entropy yield of the multi-stage screening at the corresponding parameter point.

[0125] The total entropy yield of multi-stage screening at each parameter point is integrated along the parameter point sequence to obtain the cumulative total entropy yield of multi-stage screening corresponding to each parameter point sequence. The parameter point sequence that minimizes the cumulative total entropy yield of multi-stage screening while satisfying the inter-stage material particle size distribution transmission constraint is selected as the optimal reference path.

[0126] The optimal reference path is discretized into a parameter sequence, which includes the vibration frequency sequence, amplitude sequence, vibration direction angle sequence, and feed rate sequence from the current working point to the target working point.

[0127] The analysis examines the discrepancy between the actual total entropy yield and the theoretical total entropy yield during the multi-stage sieving process, including:

[0128] The real-time input power of the drive motors of each level of vibrating screen and the real-time output of each belt scale are obtained and correlated to obtain the actual total entropy yield of multi-stage screening.

[0129] The real-time input power is collected by smart meters installed on the input side of the drive motors of each level of vibrating screen. It refers to the actual active power consumed by the drive motors of each level of vibrating screen at the current sampling time, reflecting the rate at which the vibrating screen consumes electrical energy to drive the screen frame to vibrate under the current control parameters and material load.

[0130] Real-time finished product output is obtained by a belt scale installed on the finished product conveyor belt at the outlet of each level of vibrating screen. It represents the quality of qualified finished quartz sand material discharged by each level of vibrating screen per unit time at the current sampling time, and reflects the actual screening output capacity of each vibrating screen under the current control parameters.

[0131] The real-time input power and real-time finished product output of each level of vibrating screen are summed to obtain the total input power and total finished product output of the multi-stage screening production line at the current sampling time. The unit output energy consumption at the sampling time is obtained by dividing the total input power by the total finished product output. The unit output energy consumption is divided by the absolute ambient temperature to obtain the actual total entropy yield of the multi-stage screening at the sampling time.

[0132] The theoretical total entropy yield obtained from the analysis scale entropy yield generation mechanism is compared with the actual total entropy yield from multi-stage sieving, and the multi-scale entropy yield generation mechanism is updated based on the deviation results.

[0133] The closed-loop correction module is used to analyze the deviation between the actual total entropy yield and the theoretical total entropy yield during the evolution process, update the multi-scale entropy yield generation mechanism, and use the updated multi-scale entropy yield generation mechanism as the optimization target to perform closed-loop tracking of the optimal reference path.

[0134] In actual production, the quartz sand raw materials come from different mining areas or different mining faces, and there are batch-to-batch differences in particle shape, surface roughness, and hardness distribution. By monitoring the deviation between the theoretical total entropy yield and the actual total entropy yield in real time, it is determined whether the collision recovery coefficient and rolling friction coefficient of the currently used quartz sand particles can still accurately describe the actual characteristics of the current batch of quartz sand material. The results of this determination are then used to make corrections so that the multi-scale entropy yield generation mechanism can continuously adapt to changes in actual working conditions.

[0135] The multi-scale entropy yield generation mechanism is updated based on the bias results, including:

[0136] The absolute value of the deviation between the theoretical total entropy yield and the actual total entropy yield of multi-stage screening is analyzed and compared with a preset deviation judgment threshold. The theoretical total entropy yield is calculated based on the vibration frequency, amplitude, vibration direction angle, and feed rate at the current operating point using a multi-scale entropy yield generation mechanism. Deviation = Theoretical total entropy yield - Actual total entropy yield of multi-stage screening.

[0137] It should be explained that the deviation judgment threshold is predetermined based on the statistical distribution of the deviation between the theoretical total entropy yield and the actual total entropy yield of the multi-stage screening production line under historical normal operating conditions, such as taking three times the standard deviation of the historical deviation sequence.

[0138] If the absolute value of the deviation is greater than the offset judgment threshold, an offset is determined to exist, and the collision recovery coefficient and rolling friction coefficient of the quartz sand particles are corrected according to the direction and amplitude of the deviation.

[0139] If there is an offset, it means that the collision recovery coefficient and rolling friction coefficient of the current quartz sand particles can no longer accurately describe the actual characteristics of the current batch of quartz sand material, and need to be corrected.

[0140] The specific method of correction is as follows:

[0141] If the deviation is positive, it indicates that the values ​​of the collision recovery coefficient and rolling friction coefficient of the quartz sand particles used are too large, resulting in excessively high interparticle dissipated energy calculated by the particle-scale entropy yield sub-mechanism. In this case, decrease the values ​​of the collision recovery coefficient and rolling friction coefficient of the quartz sand particles. If the deviation is negative, it indicates that the values ​​of the collision recovery coefficient and rolling friction coefficient of the quartz sand particles used are too small, resulting in excessively low interparticle dissipated energy calculated by the particle-scale entropy yield sub-mechanism. In this case, increase the values ​​of the collision recovery coefficient and rolling friction coefficient of the quartz sand particles.

[0142] The correction amount for the collision recovery coefficient and rolling friction coefficient of quartz sand particles is set to be proportional to the absolute value of the deviation. The collision recovery coefficient and rolling friction coefficient of quartz sand particles are corrected online through iterative adjustment. After each iteration, the particle size entropy yield sub-mechanism is updated using the corrected collision recovery coefficient and rolling friction coefficient of quartz sand particles, and the absolute value of the deviation between the theoretical total entropy yield and the actual total entropy yield of multi-stage screening is recalculated until it is determined that there is no deviation. Then, the iterative correction is stopped and the current coefficient value remains unchanged.

[0143] If the absolute value of the deviation is less than or equal to the offset judgment threshold, it is determined that there is no offset, and the multi-scale entropy yield generation mechanism remains unchanged. This indicates that the collision recovery coefficient and rolling friction coefficient of the current quartz sand particles can still accurately describe the actual characteristics of the current batch of quartz sand material.

[0144] Closed-loop tracking of the optimal reference path includes:

[0145] Obtain the target output for the current production task. The target output refers to the total mass of qualified quartz sand that the multi-stage screening production line needs to produce per unit time; the finished product particle size qualification rate threshold refers to the lower limit of the percentage of the mass of finished products discharged from the last vibrating screen whose particle size falls within the target finished product specification range, relative to the total mass of the finished products.

[0146] Using the updated multi-scale entropy yield generation mechanism as the optimization objective, in each rolling optimization time domain, we solve for the optimal parameter sequence that minimizes the cumulative value of the total entropy yield of multi-stage screening in the prediction time domain and satisfies that the finished product output is not lower than the target output.

[0147] The rolling optimization time domain refers to the time range during which the controller predicts and optimizes forward in each sampling period during the closed-loop rolling optimization control process. The rolling optimization time domain is jointly defined by the prediction time domain length (the time span by which the controller extends into the future from the current sampling moment) and the control time domain length (the time length for which independent control increments need to be solved and generated).

[0148] The prediction time domain is divided into multiple sampling periods, and a search is performed in the four-dimensional control parameter space to generate several candidate control parameter sequences. For each candidate control parameter sequence, the updated multi-scale entropy yield generation mechanism is invoked to obtain the multi-level screening instantaneous total entropy yield for each sampling period. The multi-level screening instantaneous total entropy yield for each sampling period in the prediction time domain is accumulated according to the sampling period to obtain the cumulative value of the multi-level screening total entropy yield corresponding to the candidate parameter sequence.

[0149] Based on the candidate parameter sequence, the screening efficiency correlation is used to find the total screening efficiency under each control parameter. The feeding speed value of each stage of vibrating screen is multiplied by the corresponding total screening efficiency to obtain the finished product output of each stage of vibrating screen at that sampling time. The finished product output of each stage of vibrating screen at each sampling time within the prediction time domain is accumulated and divided by the prediction time domain length to obtain the finished product output. The constraints are that the finished product output is not lower than the target output and the finished product particle size qualification rate is not lower than the qualification rate threshold. The candidate parameter sequence that satisfies the constraints and minimizes the cumulative value of the total entropy yield of multi-stage screening is selected as the optimal parameter sequence.

[0150] The screening efficiency correlation table refers to a data structure that is pre-established through discrete element simulation and stored in the controller. It is used to describe the quantitative correspondence between the control parameters of the vibrating screen and the screening results. The screening efficiency correlation table uses vibration frequency, amplitude, vibration direction angle, screen aperture and feeding speed as input indices, and the classification efficiency of each particle size range and the total screening efficiency as output values.

[0151] The first-step control increment from the optimal parameter sequence is superimposed onto the vibration frequency, amplitude, vibration direction angle, and feed rate of each stage of the vibrating screen, and this process is continuously refined. The first-step control increment refers to the difference between the control parameter value in the optimal parameter sequence corresponding to the first sampling period after the current sampling time and the actual control parameter value executed at the current sampling time. The first-step control increment includes the first vibration frequency increment, amplitude increment, vibration direction angle increment, and feed rate increment.

[0152] It needs to be explained that by algebraically superimposing the control increment in the first step with the vibration frequency, amplitude, vibration direction angle, and feeding speed of each level of vibrating screen, a dynamic control link based on real-time feedback correction is established. This allows the controller to use the measured current operating point of each level of vibrating screen as the starting boundary for state space optimization in each sampling cycle. This not only enables smooth evolution of control output by reducing the amplitude of single-step actions, but also allows for dynamic adjustment of the transition rate and evolution direction of parameters through a rolling optimization mechanism to address uncertain disturbances such as load fluctuations and structural deformations under actual working conditions. This ensures the robustness and anti-interference capability of the closed-loop control path during continuous evolution.

[0153] The current operating point is updated based on the real-time response of each level of the vibrating screen, and closed-loop rolling optimization control is executed until the closed-loop tracking of the optimal reference path is completed. Real-time response refers to the feedback value of the actual operating status of each level of the vibrating screen after the first step control increment is superimposed on the current control parameters and executed in each sampling cycle. This includes the actual vibration frequency, actual amplitude, actual vibration direction angle, and actual feeding speed.

[0154] In this embodiment, by applying bidirectional frequency sweep excitation to each level of vibrating screen during the low load window period, the resonant frequency and maximum amplitude of the upward and downward paths are extracted to form a nonlinear hysteresis feature vector. Based on this, the bistable frequency range is defined and the amplitude judgment boundary of the high and low amplitude stable branches is defined. The process parameters of each finished product specification are associated with their respective stable branches and stored as a process switching reference table. This realizes the explicit quantitative expression of the nonlinear hysteresis characteristics in the quartz sand screening process, enabling accurate identification of the branch to which any working point belongs during the screening process. This fundamentally avoids the branch misjudgment caused by the traditional linear strategy ignoring the hysteresis characteristics, and provides a reliable dynamic perception basis for subsequent transition decisions.

[0155] By analyzing process switching commands, the system achieves optimal global energy consumption control during process switching through phased collaboration of branch transition control and minimum entropy production path planning. When it is determined that the current operating point and the target operating point belong to different stable branches, a nonlinear resonant transition is triggered. The system is driven to cross to the target branch with low energy consumption by combining the crossing frequency with transient excitation force pulses. This solves the problem of traditional control passively staying in low-energy-efficiency branches and continuously high energy consumption. After the transition is completed, the optimal reference path that minimizes the total entropy production rate of multi-stage screening along the path is planned in the four-dimensional control parameter space. This expands the energy consumption optimization dimension from the steady-state operating point to the entire dynamic transition process. The transition-cruise collaborative architecture formed by the organic integration of the two achieves optimal global energy consumption control during large-scale process switching, significantly improving the operating energy efficiency of multi-stage quartz sand screening and the process adaptability to different finished product specifications. The closed-loop correction module monitors and corrects the multi-scale entropy production rate generation mechanism in real time, ensuring that the actual operating trajectory always closely follows the optimal reference path, further consolidating the energy-saving effect and operational stability of the process switching process.

[0156] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed in this invention can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0157] In the several embodiments provided by this invention, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only one method, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0158] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention.

[0159] 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.

Claims

1. A system for minimizing energy consumption in multi-stage screening of quartz sand, characterized in that, The energy-saving system for multi-stage screening of quartz sand includes: The feature identification module is used to obtain the nonlinear hysteresis feature vectors of each level of vibrating screen during the low load window period, divide the amplitude range of the high amplitude stable branch and the low amplitude stable branch, and form a process switching reference table. The entropy production mechanism establishment module is used to construct a multi-scale entropy production rate generation mechanism for vibrating screens at all levels within a four-dimensional control parameter space. The branch transition module is used to parse the current operating point when the process switching command is executed, extract the target operating point and its stable branch corresponding to the current operating point from the process switching reference table, and trigger a nonlinear resonance transition if it is determined that the current operating point and the target operating point do not belong to the same stable branch. The optimal path planning module is used to call the multi-scale entropy yield generation mechanism when the resonance transition is completed or the same stable branch is determined. It plans the optimal reference path in the four-dimensional control parameter space that minimizes the total entropy yield of the multi-stage sieving along the path integral and forms a parameter sequence that evolves towards the target working point. The closed-loop correction module is used to analyze the deviation between the actual total entropy yield and the theoretical total entropy yield during the evolution process, update the multi-scale entropy yield generation mechanism, and use the updated multi-scale entropy yield generation mechanism as the optimization target to perform closed-loop tracking of the optimal reference path.

2. The energy-saving system for multi-stage screening of quartz sand according to claim 1, characterized in that, The acquisition of the nonlinear hysteresis characteristic vectors of each level of vibrating screen during the low load window period includes: An upward sweep frequency excitation signal with linearly increasing frequency and a downward sweep frequency excitation signal with linearly decreasing frequency are applied to each level of vibrating screen during the low load window period. The vibration acceleration signals of each level of vibrating screen during the sweep frequency are collected synchronously and converted into time and frequency to obtain the upward amplitude-frequency characteristic curve and the downward amplitude-frequency characteristic curve of each level of vibrating screen. The peak amplitude and corresponding frequency are extracted from the uplink amplitude-frequency characteristic curve and used as the maximum amplitude and resonant frequency of the uplink path, respectively; the amplitude value and frequency before the amplitude drops sharply in the downlink amplitude-frequency characteristic curve are identified and used as the maximum amplitude and resonant frequency of the downlink path, respectively. The maximum amplitude of the upward path, the resonant frequency of the upward path, the maximum amplitude of the downward path, and the resonant frequency of the downward path are combined to form the nonlinear hysteresis characteristic vector of each level of the vibrating screen.

3. The energy-saving system for multi-stage screening of quartz sand according to claim 2, characterized in that, The process switching reference table includes: The bistable frequency range is defined by the resonant frequencies of the uplink and downlink paths; Within the bistable frequency range, the maximum amplitudes of the downlink and uplink paths are used as the lower and upper bounds for determining the amplitude range of the high-amplitude stable branch, respectively. The amplitude value at which the amplitude suddenly drops and then re-stabilizes in the downlink amplitude-frequency characteristic curve is extracted and used, along with the minimum allowable stable amplitude, as the upper and lower bounds for determining the amplitude range of the low-amplitude stable branch, respectively. A mapping relationship database is established based on historical quartz sand screening data; the mapping relationship database records the mapping relationship between the finished quartz sand specifications and the corresponding four-dimensional control parameters; The amplitude of each finished product specification is compared with the amplitude range of each stable branch. The stable branch obtained from the comparison is associated with the corresponding four-dimensional control parameters to form a process switching reference table.

4. The energy-saving system for multi-stage screening of quartz sand according to claim 2, characterized in that, The mechanism for constructing a multi-scale entropy yield generation method for vibrating screens at various levels includes: Within a four-dimensional control parameter space consisting of frequency, amplitude, vibration direction angle and feeding speed, a multi-scale entropy yield generation mechanism is constructed for each level of vibrating screen, including particle-scale entropy yield sub-mechanisms and screen-scale entropy yield sub-mechanisms. The particle-scale entropy yield sub-mechanism is used to obtain the particle-scale entropy yield, which characterizes the inelastic collision and friction energy dissipation of quartz sand particles, by logical deduction based on the measured collision recovery coefficient and rolling friction coefficient of quartz sand particles, combined with the vibration direction angle and feeding speed. The screen-scale entropy yield sub-mechanism is used to analyze the loss factor of screen material measured by vibrating screens at each level, combined with the metal structural stiffness, vibration frequency and amplitude of each level of vibrating screen, to obtain the screen-scale entropy yield that characterizes the internal friction and elastic hysteresis dissipation of screen material at each level of vibrating screen.

5. The energy-saving system for multi-stage screening of quartz sand according to claim 3, characterized in that, The extraction of the target working point and its stable branch corresponding to the current working point includes: The received process switching command is parsed to obtain the target finished product specifications; the vibration frequency, amplitude, vibration direction angle and feeding speed of each level of vibrating screen are taken as the current working point; based on the vibration frequency and amplitude of the current working point, combined with the nonlinear hysteresis characteristic vector, the stable branch to which the current working point belongs is determined. Retrieve the stable branch corresponding to the target finished product specification from the process switching reference table, as well as the target four-dimensional control parameters that serve as the corresponding target working point; determine whether the stable branch to which the current working point and the target working point belong are the same. If they are not the same, trigger a nonlinear resonant transition.

6. The energy-saving system for multi-stage screening of quartz sand according to claim 5, characterized in that, The triggering of the nonlinear resonant transition includes: Calculate the average value of the resonance frequencies of the upward and downward paths of each level of vibrating screen as the crossing frequency; adjust the excitation frequency from the vibration frequency of the current working point to the crossing frequency according to the first preset rate of change, and apply a transient excitation force pulse to the excitation motor synchronously when it is determined that the excitation frequency crosses the crossing frequency. After the transient excitation force pulse is applied, the response amplitude of each level of the vibrating screen is monitored; when each response amplitude jumps to the amplitude range of the stable branch to which the target working point belongs, the excitation frequency is adjusted to the vibration frequency corresponding to the target working point with the second preset change rate, so as to complete the resonance transition and frequency alignment from the current working point to the stable branch to which the target working point belongs.

7. The energy-saving system for multi-stage screening of quartz sand according to claim 6, characterized in that, The plan identifies the optimal reference path that minimizes the integral of the total entropy yield of the multi-stage screening process along the path, including: Establish inter-stage material particle size distribution transmission constraints; under the constraint of inter-stage material particle size distribution transmission, the multi-scale entropy yield obtained by the multi-scale entropy yield generation mechanism is the cost target. The current working point is taken as the planning start point and the target working point is taken as the planning end point. Search and plan in the four-dimensional control parameter space to obtain the optimal reference path that minimizes the total entropy yield of multi-stage screening along the path. The optimal reference path is discretized into a parameter sequence, which includes the vibration frequency sequence, amplitude sequence, vibration direction angle sequence, and feed rate sequence from the current working point to the target working point.

8. The energy-saving system for multi-stage screening of quartz sand according to claim 7, characterized in that, The deviations between the actual total entropy yield and the theoretical total entropy yield during the multi-stage sieving process described in the analysis evolution include: The real-time input power of the drive motors of each level of vibrating screen and the real-time finished product output fed back by each belt scale are obtained and correlated to obtain the actual total entropy yield of multi-stage screening. The theoretical total entropy yield obtained from the analysis scale entropy yield generation mechanism is compared with the actual total entropy yield from multi-stage sieving, and the multi-scale entropy yield generation mechanism is updated based on the deviation results.

9. The energy-saving system for multi-stage screening of quartz sand according to claim 8, characterized in that, The update of the multi-scale entropy yield generation mechanism based on the deviation results includes: The absolute value of the deviation between the theoretical total entropy yield and the actual total entropy yield of multi-stage screening is analyzed and compared with the preset deviation judgment threshold. If the absolute value of the deviation is greater than the deviation judgment threshold, it is determined that there is a deviation. The collision recovery coefficient and rolling friction coefficient of quartz sand particles are corrected according to the direction and amplitude of the deviation. If the absolute value of the deviation is less than or equal to the offset determination threshold, it is determined that there is no offset, and the multi-scale entropy yield generation mechanism remains unchanged.

10. The energy-saving system for multi-stage screening of quartz sand according to claim 9, characterized in that, The closed-loop tracking of the optimal reference path includes: Obtain the target output of the current production task; with the updated multi-scale entropy yield generation mechanism as the optimization objective, in each rolling optimization time domain, solve for the optimal parameter sequence that minimizes the cumulative value of the total entropy yield of multi-stage screening in the prediction time domain and satisfies that the finished product output is not lower than the target output. The first step control increment in the optimal parameter sequence is superimposed on the vibration frequency, amplitude, vibration direction angle and feeding speed of the current level of vibrating screen and continuously evolved; the current working point is updated according to the real-time response of each level of vibrating screen, and closed-loop rolling optimization control is executed until the closed-loop tracking of the optimal reference path is completed.

Images