Automatic cleaning control system for crusher assembly line

Abstract

The invention relates to the technical field of industrial automation control, and discloses a crusher assembly line automatic cleaning control system which comprises a pretreatment module, a crushing control module, a dry cleaning control module, a logistics transmission module, a sensor monitoring module, a central control module and the like. The pretreatment module automatically discharges oil and transfers waste engine oil; the crushing control module adjusts crushing parameters based on the pressure sensor and generates particle size distribution data; the dry cleaning control module dynamically adjusts a chemical agent and airflow according to the working state; the logistics transmission module optimizes a material transfer path and priority; the sensor monitoring module acquires multi-dimensional data in real time and gives an alarm; the central control module coordinates the time sequence of each module and allocates resources. In addition, a waste recovery and data storage module is further arranged. According to the system, automation and intelligence of the cleaning process of the crusher assembly line are achieved, the cleaning efficiency and quality are improved, stable operation of the system is guaranteed, and the system has the advantages of resource recycling, data safety management and the like.

Description

Technical Field

[0001] The present invention relates to the technical field of industrial automation control, and more specifically to an automatic cleaning control system for a crusher production line. Background Art

[0002] In modern industrial production, crusher production lines are widely used in the crushing treatment of various materials. Especially when dealing with oil-contaminated parts such as engine components, the cleaning work is crucial. Most traditional cleaning methods for crusher production lines rely on manual operation, which has many drawbacks and severely restricts production efficiency and cleaning quality.

[0003] The oil draining operation of manually cleaning engine components is inefficient and incomplete, and it is difficult to accurately control the remaining amount of engine oil. This not only causes waste of resources but also may pollute the environment due to improper treatment of waste engine oil. In the crushing process, manual workers cannot real-time monitor the pressure changes in the crushing chamber, and it is difficult to timely adjust the rotational speed of the crushing cutter head and the feeding speed according to the material characteristics, resulting in uneven particle size distribution of the crushed material and affecting subsequent processing and product quality.

[0004] The dry cleaning process also lacks effective automatic control means. The working conditions of the dry cleaning chamber such as temperature and vibration are difficult to monitor in real time. The chemical agent dosing and air flow circulation depend entirely on manual experience and cannot be dynamically adjusted according to the preset dry cleaning formula parameters. This makes the use of chemical agents unreasonable, which not only increases costs but also may damage the material due to incomplete cleaning or over-cleaning. At the same time, uneven air flow distribution leads to poor cleaning uniformity and is difficult to meet the high-quality cleaning requirements.

[0005] In terms of material handling, the material transfer path planning between the loader and the electromagnetic chuck lacks systematicness and scientificity. Relying on manual scheduling, it is impossible to optimize the transfer path and priority in real time according to the stacking height of the crushed material and the capacity of the dry cleaning chamber, resulting in material backlog or equipment idleness and reducing the operating efficiency of the entire production line. Moreover, it is difficult for manual operation to comprehensively and real-time monitor multi-dimensional data during the cleaning process, such as oil concentration, metal debris situation, and environmental temperature and humidity. Once an abnormal situation occurs, it cannot be discovered and processed in time, affecting the continuity and stability of production.

[0006] With the continuous development of industrial automation technology, the market's requirements for the automation, intelligence, and high efficiency of crusher production line cleaning systems are increasing day by day. The existing cleaning systems can no longer meet these needs, and there is an urgent need for an automatic cleaning control system for a crusher production line that can achieve full-process automatic control, accurate monitoring, and efficient operation to improve production efficiency, reduce costs, improve product quality, and reduce the impact on the environment. Summary of the Invention

[0007] The purpose of the present invention is to provide an automatic cleaning control system for a crusher production line to solve the problems raised in the above background art.

[0008] To achieve the above object, the present invention provides the following technical solution: An automatic cleaning control system for a crusher production line, the system comprising:

[0009] A pretreatment module for draining the engine components, detecting the residual amount of engine oil through an oil level sensor, controlling an oil suction device to transfer the waste engine oil to a temporary storage area, and generating a waste oil transfer instruction;

[0010] A crushing control module for receiving the completion signal from the pretreatment module, starting the crusher to crush the oil-contaminated parts, and based on the pressure sensor in the crushing chamber, adjusting the rotation speed of the crushing cutter head and the feeding speed in real time, and generating crushing material particle size distribution data;

[0011] A dry cleaning control module for conveying the crushed material to a dry cleaning device, monitoring the working state of the dry cleaning chamber through a temperature sensor and a vibration sensor, and dynamically adjusting the chemical agent feeding rate and the air flow circulation frequency based on preset dry cleaning formula parameters;

[0012] A logistics transmission module for coordinating the material transfer paths of a loader and an electromagnetic chuck, generating a material scheduling priority queue according to the stacking height of the crushed material and the capacity of the dry cleaning chamber, and optimizing the transmission efficiency;

[0013] A sensor monitoring module, including an oil pollution concentration detection unit, a metal debris identification unit, and an environmental temperature and humidity acquisition unit, for obtaining multi-dimensional data during the cleaning process in real time and generating an abnormal alarm signal;

[0014] A central control module for integrating the data streams of the pretreatment module, the crushing control module, the dry cleaning control module, the logistics transmission module, and the sensor monitoring module, coordinating the execution timing of each module through a rule-based state machine model, and allocating system resources based on a dynamic weight scheduling algorithm.

[0015] Preferably, the crushing control module further includes:

[0016] A particle size feedback unit for measuring the particle size distribution of the crushed material in real time through a laser diffractometer, and generating a cutter head gap adjustment instruction according to a preset particle size threshold;

[0017] A load balancing unit for dynamically allocating the load ratio of multiple groups of crushing cutter heads based on the motor current signal and the vibration spectrum analysis result to prevent local overload. Specifically, the load distribution weight of the cutter head is calculated through the following formula:

[0018]

[0019] where w i is the load weight of the i-th group of cutter heads, and I iis the amplitude of the motor current signal of the i-th cutter head, F i is the main frequency amplitude vector of the vibration spectrum of the i-th cutter head, and n is the cutter head group index.

[0020] Preferably, the dry cleaning control module further includes:

[0021] A chemical agent ratio optimization unit for calculating the mapping relationship between the chemical agent dosage and the oil stain removal rate through a non-linear regression model based on the data of the oil stain concentration detection unit, and dynamically adjusting the ratio parameters;

[0022] An air flow path planning unit for optimizing the air flow distribution by using a vortex generation algorithm based on the three-dimensional flow field simulation data in the dry cleaning chamber to improve the cleaning uniformity, and the vortex generation algorithm satisfies:

[0023]

[0024] Among them, V s is the air flow velocity at the vortex center, α is the vortex intensity coefficient, ΔP is the pressure difference inside and outside the dry cleaning chamber, and ρ is the gas density.

[0025] Preferably, the logistics transmission module further includes:

[0026] A path planning unit for constructing a material transfer path by using a grid map based on the magnetic force coverage range of the electromagnetic chuck and the kinematic model of the loader, and real-time correcting the path through a collision avoidance constraint algorithm;

[0027] A priority dynamic adjustment unit for updating the priority weight of the material scheduling queue by using a sliding window algorithm according to the crushing material accumulation rate and the idle capacity of the dry cleaning chamber, and the priority weight update formula is:

[0028]

[0029] Among them, W j is the priority weight of the j-th queue, R j is the crushing material accumulation rate of the j-th queue, C j is the current dry cleaning chamber capacity of the j-th queue, C max is the maximum capacity of the dry cleaning chamber, and β is the normalization coefficient.

[0030] Preferably, the oil stain concentration detection unit in the sensor monitoring module adopts infrared spectroscopy analysis technology to calculate the oil stain residue through the absorption rate of the characteristic wavelength, and combines the Kalman filtering algorithm to eliminate the environmental noise interference.

[0031] Preferably, the dynamic weight scheduling algorithm in the central control module includes:

[0032] A resource occupancy evaluation unit, which is used to count the real-time computing load and communication delay of each module and generate a resource occupancy matrix;

[0033] A weight allocation unit, which is used to determine the priority coefficient of each module based on the matrix singular value decomposition result and optimize the weight allocation strategy by the gradient descent method.

[0034] Preferably, the system further includes:

[0035] A waste recycling module, which is used to separate the metal and non-metal components in the crushed material by a magnetic separator, adjust the magnetic separation intensity based on the data of the metal debris identification unit, and optimize the separation efficiency through closed-loop feedback control.

[0036] Preferably, the waste recycling module further includes:

[0037] A hysteresis compensation unit, which is used to predict the hysteresis effect by using a polynomial fitting model according to the temperature change curve and historical magnetic flux data of the magnetic separator, and dynamically correct the magnetic separation parameters. The magnetic flux prediction model is:

[0038] Φ(T) = a0 + a1T + a2T 2

[0039] where Φ(T) is the predicted magnetic flux at temperature T, and a0, a1, a2 are fitting coefficients.

[0040] Preferably, the system further includes:

[0041] A data storage module, which is used to encrypt and store the original data of the sensor monitoring module and the decision log of the central control module according to the time stamp, and desensitize the sensitive data by the differential privacy algorithm.

[0042] Preferably, the data storage module further includes:

[0043] A data compression unit, which is used to perform dimensionality reduction processing on the high-frequency sampling data by using the wavelet transform algorithm and achieve lossless compression through Huffman coding.

[0044] Compared with the prior art, the beneficial effects of the present invention are:

[0045] The pre - treatment module accurately detects the remaining amount of engine oil through an oil level sensor, controls the oil suction device to automatically transfer waste engine oil, ensuring efficient and thorough oil draining operations. Compared with manual oil draining, it greatly shortens the oil draining time, avoids the impact of remaining engine oil on subsequent processes. At the same time, the generation of waste oil transfer instructions realizes seamless connection with the subsequent crushing process, improving the coherence of the entire cleaning process. The crushing control module automatically adjusts the rotation speed of the crushing cutter head and the feeding speed based on the real - time data of the pressure sensor in the crushing chamber, effectively ensuring the stability of the crushing process and improving the crushing efficiency. Meanwhile, the particle size feedback unit uses a laser diffractometer to measure the particle size distribution in real - time and adjusts the cutter head gap to ensure that the particle size of the crushed material meets the preset requirements, enhancing the consistency of product quality. The load - balancing unit dynamically distributes the cutter head load through motor current signals and vibration spectrum analysis, preventing local overload, extending the service life of the equipment, and reducing the equipment maintenance cost.

[0046] The dry - cleaning control module uses temperature sensors and vibration sensors to monitor the working state of the dry - cleaning chamber in real - time, and dynamically adjusts the chemical agent feeding rate and the air - flow circulation frequency according to the preset dry - cleaning formula parameters. The chemical agent ratio optimization unit accurately adjusts the chemical agent feeding amount through a non - linear regression model based on the oil stain concentration detection data, reducing chemical agent waste while ensuring the cleaning effect and lowering production costs. The air - flow path planning unit uses a vortex generation algorithm to optimize the air - flow distribution, significantly improving the cleaning uniformity and ensuring comprehensive and effective cleaning of the materials.

[0047] The logistics transmission module coordinates the material transfer paths of the loader and the electromagnetic chuck, generates a material scheduling priority queue based on the stacking height of the crushed material and the capacity of the dry - cleaning chamber, realizing intelligent and efficient material transfer. The path planning unit constructs the transfer path based on the magnetic force coverage range of the electromagnetic chuck and the kinematic model of the loader, and corrects it in real - time through a collision - avoidance constraint algorithm, improving the transfer safety and accuracy. The priority dynamic adjustment unit uses a sliding window algorithm to update the priority weights, ensuring timely material transfer, avoiding material backlog and equipment idleness, and greatly improving the production efficiency of the entire production line.

[0048] The sensor monitoring module uses infrared spectroscopy analysis technology combined with the Kalman filtering algorithm to detect the oil stain concentration, which is accurate and reliable, effectively eliminating environmental noise interference. The metal debris identification unit and the environmental temperature and humidity acquisition unit obtain multi - dimensional data in real - time. Once an abnormal situation occurs, an alarm signal is immediately generated, facilitating operators to take timely measures, ensuring the safety and stability of the production process, and reducing production interruptions and product losses caused by abnormal situations.

[0049] The central control module coordinates the execution timing of each module through a rule-based state machine model to ensure the orderly operation of the system. The dynamic weight scheduling algorithm reasonably allocates system resources according to the resource occupancy of each module, improving the overall performance of the system. The data storage module encrypts and desensitizes the sensor data and decision logs, ensuring data security. At the same time, wavelet transform and Huffman coding are used to compress the data, saving storage space and facilitating the long-term preservation and management of data. The waste recycling module effectively separates the metal and non-metal components in the crushed materials through a magnetic separator, adjusts the magnetic separation intensity based on the data of the metal debris identification unit, and optimizes the separation efficiency using closed-loop feedback control, realizing the recycling of waste, improving the resource utilization rate, and reducing environmental pollution caused by waste. Description of the Drawings

[0050] Figure 1 It is the working principle diagram of the automatic cleaning control system of the crusher production line described in the present invention;

[0051] Figure 2 It is the working flowchart of the crushing control module;

[0052] Figure 3 It is the working flowchart of the dry cleaning control module;

[0053] Figure 4 It is the working flowchart of the logistics transmission module. Detailed Embodiments

[0054] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0055] Please refer to Figures 1-4 , the present invention provides a technical solution: an automatic cleaning control system for a crusher production line, which integrates multiple functional modules to jointly realize the automatic cleaning and material handling of the crusher production line. Specifically, it includes the following modules:

[0056] Pretreatment module: Perform an oil draining operation on the engine components. The system pre-installs an oil level sensor at a suitable position of the engine components, and the oil level sensor continuously monitors the remaining amount of engine oil. When engine oil is detected, the oil suction device is controlled to start. The oil suction device is connected to the temporary storage area through a pipeline to smoothly transfer the waste engine oil. During the transfer process, the remaining amount of engine oil is continuously monitored until the remaining amount of engine oil is lower than the set threshold, completing the oil draining operation and generating a waste oil transfer instruction.

[0057] Crushing control module: After receiving the completion signal from the pretreatment module, the crushing control module starts the crusher. A pressure sensor is installed inside the crusher to monitor the pressure in the crushing chamber in real time. Through the feedback of pressure data, the rotation speed of the crushing cutter head and the feeding speed are automatically adjusted. For example, when the pressure in the crushing chamber is too high, the feeding speed is reduced, and at the same time, the rotation speed of the crushing cutter head is appropriately increased to ensure the smooth progress of the crushing process. During the crushing process, data is continuously collected to generate particle size distribution data of the crushed material.

[0058] Dry cleaning control module: The material after crushing is conveyed to the dry cleaning equipment by the conveying device. A temperature sensor and a vibration sensor are installed inside the dry cleaning chamber to monitor the working state of the dry cleaning chamber in real time. According to the preset dry cleaning formula parameters, the system dynamically adjusts the chemical agent feeding rate and the air flow circulation frequency. For example, when the temperature in the dry cleaning chamber is too high, the chemical agent feeding rate is appropriately reduced, and at the same time, the air flow circulation frequency is adjusted to ensure that the dry cleaning process is carried out under suitable conditions.

[0059] Logistics transmission module: Coordinate the material transfer paths of the loader and the electromagnetic chuck. Multiple sensors are set in the working area to monitor the stacking height of the crushed material and the capacity of the dry cleaning chamber in real time. According to this data, a material scheduling priority queue is generated to optimize the material transmission efficiency. For example, when the stacking height of the crushed material in a certain area is relatively high and the corresponding dry cleaning chamber has a relatively large capacity, the loader and the electromagnetic chuck are preferentially arranged to transfer the material in this area.

[0060] Sensor monitoring module: It includes an oil concentration detection unit, a metal chip identification unit, and an environmental temperature and humidity acquisition unit. The oil concentration detection unit detects the oil concentration in the cleaning process in real time, the metal chip identification unit identifies the metal chips in the material, and the environmental temperature and humidity acquisition unit acquires the environmental temperature and humidity data. When the monitored data exceeds the preset threshold, an abnormal alarm signal is generated to notify the operator to handle it in time.

[0061] Central control module: Integrate the data streams of the pretreatment module, the crushing control module, the dry cleaning control module, the logistics transmission module, and the sensor monitoring module. Through a rule-based state machine model, strictly coordinate the execution timing of each module. For example, ensure that the crushing control module starts only after the pretreatment module is completed. At the same time, based on a dynamic weight scheduling algorithm, reasonably allocate system resources to ensure the efficient and stable operation of the system.

[0062] The present invention will be further described below in conjunction with Embodiments 1 to 6:

[0063] Embodiment 1:

[0064] In the crushing control module, the particle size feedback unit measures the particle size distribution of the crushed material in real time through a laser diffraction instrument. The laser diffraction instrument is installed near the discharge port of the crusher and can detect the discharged crushed material in real time. It irradiates the crushed material with a laser beam. According to the principle of light diffraction, particles of different particle sizes will produce different diffraction patterns. By analyzing the diffraction patterns, the particle size distribution of the crushed material is accurately calculated.

[0065] The system pre-sets a particle size threshold. When the particle size distribution measured by the laser diffraction instrument exceeds the pre-set particle size threshold, the particle size feedback unit generates a cutter head gap adjustment instruction. This instruction is transmitted to the cutter head adjustment mechanism of the crusher, and the cutter head adjustment mechanism accurately adjusts the gap between the crushing cutter heads according to the instruction. For example, if the measured particle size of the crushed material is too large, it indicates that the cutter head gap may be too large. The cutter head adjustment mechanism will reduce the cutter head gap so that the subsequent crushed material can meet finer particle size requirements.

[0066] The load balancing unit dynamically distributes the load ratios of multiple groups of crushing cutter heads based on the motor current signal and the vibration spectrum analysis result to prevent local overload. Current sensors are installed on the motors of each group of crushing cutter heads to collect the amplitude I of the motor current signal in real time i At the same time, vibration sensors are installed at key parts of the crusher. By performing spectrum analysis on the vibration signals, the main frequency amplitude vector F of the vibration spectrum is obtained i .

[0067] According to the formula Calculate the load distribution weight w of the cutter head i . For example, assume there are 3 groups of cutter heads. At a certain moment, the amplitude I1 of the motor current signal of the first group of cutter heads is 2A, and the modulus length ‖F1‖ of the main frequency amplitude vector F1 of the vibration spectrum is calculated to be 3; for the second group of cutter heads, I2 = 1.5A, ‖F2‖ = 2; for the third group of cutter heads, I3 = 2.5A, ‖F3‖ = 2.5. Then first calculate the denominator The load weight of the first group of cutter heads The load weight of the second group of cutter heads The load weight of the third group of cutter heads According to the calculated load weights, the system automatically adjusts the working parameters of each group of cutter heads to make the loads of each cutter head more balanced and avoid local overload.

[0068] Example 2:

[0069] In the dry cleaning control module, the chemical agent ratio optimization unit calculates the mapping relationship between the chemical agent dosage and the oil stain removal rate through a non-linear regression model based on the data of the oil stain concentration detection unit, and dynamically adjusts the ratio parameters. The oil stain concentration detection unit uses specific sensor technology to detect the oil stain concentration on the surface of the material in real time during the dry cleaning process.

[0070] The system pre-collects a large amount of data on the chemical agent dosage and oil stain removal rate under different oil stain concentrations, and uses this data to train a non-linear regression model. For example, a polynomial regression model y = a0 + a1x + a2x 2 +…+ a n x n is adopted, where y represents the oil stain removal rate, x represents the chemical agent dosage, and a0, a1, …, a n are model parameters. Through optimization algorithms such as the least squares method, the model parameters are determined to enable the model to accurately reflect the relationship between the chemical agent dosage and the oil stain removal rate.

[0071] When the oil stain concentration changes during the dry cleaning process, the chemical agent ratio optimization unit inputs the real-time oil stain concentration data into the trained non-linear regression model, calculates the current optimal chemical agent dosage, and then adjusts the parameters of the chemical agent dosing device to achieve dynamic adjustment of the chemical agent dosage.

[0072] Based on the three-dimensional flow field simulation data in the dry cleaning chamber, the air flow path planning unit uses the vortex generation algorithm to optimize the air flow distribution and improve the cleaning uniformity. During the dry cleaning chamber design stage, the computational fluid dynamics (CFD) software is used to perform three-dimensional simulation of the air flow in the dry cleaning chamber to obtain the three-dimensional flow field data under different working conditions.

[0073] According to the vortex generation algorithm formula where V s is the air flow velocity at the vortex center, α is the vortex intensity coefficient (predetermined according to the structure and working requirements of the dry cleaning chamber), ΔP is the pressure difference inside and outside the dry cleaning chamber, and ρ is the gas density. A pressure sensor is installed inside the dry cleaning chamber to measure the pressure difference ΔP inside and outside the dry cleaning chamber in real time, and at the same time obtain the current gas density ρ.

[0074] Based on the calculated air flow velocity V s at the vortex center, the system adjusts the flow velocity and direction of the air flow by controlling equipment such as the fan to generate a specific vortex air flow inside the dry cleaning chamber, making the air flow distribution more uniform, thereby improving the cleaning uniformity of the materials. For example, when the calculated V s is small, the fan power is appropriately increased to increase the air flow velocity and enhance the vortex effect.

[0075] Example 3:

[0076] Based on the magnetic force coverage range of the electromagnetic chuck and the kinematic model of the loader, the path planning unit in the material handling and transportation module constructs the material transfer path using a rasterized map. First, the working area is rasterized, and the entire area is divided into multiple grids of equal size. According to the magnetic force coverage range of the electromagnetic chuck, the area where it can effectively adsorb materials on the grid map is determined.

[0077] Meanwhile, a kinematic model of the loader is established, considering parameters such as the turning radius of the loader and the forward and backward speeds. When planning the material transfer path, the goal is for the loader to start from the initial position, reach the material stacking area, and then transfer the material to the dry cleaning bin. The path planning algorithm starts from the grid where the loader's initial position is located, and according to the adsorption range of the electromagnetic chuck and the kinematic constraints of the loader, searches for a feasible path to the material stacking area, and then searches for a feasible path from the material stacking area to the dry cleaning bin, finally obtaining a complete material transfer path.

[0078] During the transfer process, the path is corrected in real time through the collision avoidance constraint algorithm. Multiple obstacle sensors are installed in the working area to monitor in real time whether there are obstacles. When an obstacle is detected, the collision avoidance constraint algorithm corrects the planned path according to the position and size of the obstacle. For example, if an obstacle is detected in front of the path, the algorithm will search for other feasible paths around the obstacle on the grid map to ensure that the loader and the electromagnetic chuck do not collide with the obstacle during the material transfer process.

[0079] The priority dynamic adjustment unit updates the priority weights of the material scheduling queue using the sliding window algorithm according to the crushing material stacking rate and the idle capacity of the dry cleaning bin. Sensors are installed near the crushing area and the dry cleaning bin respectively to monitor the crushing material stacking rate R j and the current capacity C of the dry cleaning bin j , and at the same time, the maximum capacity C of the dry cleaning bin is known max .

[0080] The system sets a sliding window size, for example, the window size is 5 minutes. Within each sliding window, calculate the priority weight W of each queue (corresponding to different combinations of material stacking areas and dry cleaning bins) j . According to the formula where β is the normalization coefficient (pre-set according to the actual situation to ensure that W j is within a reasonable range). For example, assume that the crushing material stacking rate R j of a certain queue is 5m 3 / h, the maximum capacity C of the dry cleaning bin max is 20m 3 , the current capacity C j is 10m 3 , and β = 0.5, then the priority weight of this queue Over time, the sliding window moves continuously, and the priority weights of each queue are updated in real time. The system adjusts the material scheduling order according to the updated priority weights, preferentially transferring materials with high priorities, and improving the overall transmission efficiency.

[0081] Example 4:

[0082] This embodiment elaborates in detail the working principle and data processing method of the oil contamination concentration detection unit in the sensor monitoring module to improve the accuracy of oil contamination concentration detection.

[0083] The oil contamination concentration detection unit in the sensor monitoring module adopts infrared spectroscopy analysis technology. An infrared spectrometer is installed at an appropriate position during the dry cleaning equipment or material transfer process. The spectrometer emits infrared light within a specific wavelength range. When the infrared light irradiates the surface of the material containing oil contamination, specific components in the oil contamination will absorb part of the infrared light.

[0084] Different types of oil contamination have different absorption rates at specific wavelengths. By measuring the absorption rates at these characteristic wavelengths, the residual amount of oil contamination can be calculated. For example, for a certain common oil contamination, it has obvious absorption peaks at wavelengths λ1, λ2, and λ3. By measuring the absorption rates A1, A2, and A3 at these three wavelengths and using a pre-established relationship model between the absorption rate and the residual amount of oil contamination (such as a linear regression model or a more complex non-linear model), the residual amount of oil contamination on the material surface can be calculated.

[0085] However, during the actual detection process, environmental noise will interfere with the detection results. To eliminate the interference of environmental noise, the oil contamination concentration detection unit combines the Kalman filtering algorithm. The Kalman filtering algorithm is an optimal estimation algorithm based on the state space model of a linear system. First, a state space model of the oil contamination concentration is established, with the oil contamination concentration as the state variable and the measured absorption rate as the observation variable.

[0086] At the initial moment, the initial estimated value and the initial error covariance of the oil contamination concentration are set. As time goes by, each time new absorption rate measurement data is received, the Kalman filtering algorithm continuously optimizes the estimation of the oil contamination concentration through two steps: prediction and update, based on the state estimation value at the previous moment and the current measurement value. The prediction step predicts the state value and the error covariance at the current moment according to the dynamic model of the system; the update step corrects the predicted value using the current measurement value to obtain a more accurate state estimation value. Through the processing of the Kalman filtering algorithm, the accuracy of oil contamination concentration detection is effectively improved, and the influence of environmental noise on the detection results is reduced.

[0087] Example 5:

[0088] The dynamic weight scheduling algorithm in the central control module includes a resource occupancy evaluation unit and a weight allocation unit. The resource occupancy evaluation unit statistics the real-time computing load and communication delay of each module and generates a resource occupancy matrix. During the operation of the system, a dedicated monitoring program is set for each module to monitor its computing load in real time, such as indicators like CPU usage rate and memory occupancy, and at the same time monitor the communication delay time between modules.

[0089] Organize the computational load and communication delay data of each module into a resource occupancy matrix. Assume that the system has m modules, and the resource occupancy matrix A is an m×2 matrix, where the element a in the first column of the i-th row i1 represents the computational load of the i-th module, and the element a in the second column of the i-th row i2 represents the communication delay of the i-th module.

[0090] The weight allocation unit determines the priority coefficients of each module based on the matrix singular value decomposition result and optimizes the weight allocation strategy through the gradient descent method. Perform singular value decomposition on the resource occupancy matrix A to obtain A = UΣV T , where U and V are orthogonal matrices, and Σ is a diagonal matrix, and the elements on the diagonal are singular values. Determine the priority coefficients of each module according to the magnitudes of the singular values. The larger the singular value, the greater the demand of the module for system resources and the higher the priority.

[0091] Optimize the weight allocation strategy through the gradient descent method. Set an objective function, such as maximizing the overall system resource utilization or minimizing the system response time. Take the weight vector w = (w1, w2, …, w m ) as the optimization variable, and the objective function is J(w). In each iteration process, calculate the gradient of the objective function with respect to the weight vector Update the weight vector according to the gradient direction, that is where η is the learning rate (predetermined according to the actual situation). Keep iterating until the objective function converges to obtain the optimal weight allocation strategy and achieve the reasonable allocation of system resources.

[0092] Example 6:

[0093] The waste recycling module separates the metal and non-metal components in the crushed material through a magnetic separator. At the feed inlet of the magnetic separator, the crushed material evenly enters the magnetic separation area. The magnetic separator generates a magnetic field, and the metal components are adsorbed on the surface of the drum of the magnetic separator under the action of the magnetic field. As the drum rotates, they are carried to one side of the discharge port, while the non-metal components directly fall under the action of gravity and other forces, realizing the preliminary separation of the metal and non-metal components.

[0094] Adjust the magnetic separation intensity based on the data of the metal debris recognition unit. The metal debris recognition unit uses image recognition or other detection technologies to detect the content and characteristics of metal debris in the crushed material in real time. When it detects a high content of metal debris or special metal components, the system automatically adjusts the current or magnetic field intensity of the magnetic separator to improve the magnetic separation effect. For example, if it detects metal debris with weak magnetism, appropriately increase the current of the magnetic separator to enhance the magnetic field intensity to ensure that these metal debris can be effectively adsorbed.

[0095] Optimize the separation efficiency through closed-loop feedback control. Detection sensors for metal and non-metal components are respectively set at the discharge port of the magnetic separator to detect the purity of the separated metal and non-metal components in real time. If the content of non-metal impurities in the metal component is too high, it indicates that the magnetic separation intensity is insufficient, and the system increases the magnetic field intensity of the magnetic separator; if the content of metal impurities in the non-metal component is too high, it indicates that the magnetic separation intensity is too high, and the system reduces the magnetic field intensity of the magnetic separator. Through continuous feedback adjustment, the optimization of separation efficiency is achieved.

[0096] The hysteresis compensation unit in the waste recycling module predicts the hysteresis effect using a polynomial fitting model based on the temperature change curve and historical magnetic flux data of the magnetic separator, and dynamically corrects the magnetic separation parameters. Install a temperature sensor on the magnetic separator to monitor the temperature T of the magnetic separator in real time. At the same time, record the historical magnetic flux data of the magnetic separator under different working conditions.

[0097] According to the formula Φ(T) = a0 + a1T + a2T 2 , using fitting methods such as the least squares method, determine the fitting coefficients a0, a1, and a2 based on the historical temperature and magnetic flux data. For example, collect the magnetic flux data of the magnetic separator at different temperatures over a period of time, substitute these data into the formula, and through fitting calculation, obtain a0 = 1.2, a1 = 0.05, and a2 = -0.001. When the temperature of the magnetic separator changes, predict the magnetic flux Φ(T) according to this model, and then dynamically correct the working parameters of the magnetic separator, such as current, magnetic field intensity, etc., to compensate for the influence brought by the hysteresis effect.

[0098] The data storage module encrypts and stores the original data of the sensor monitoring module and the decision-making log of the central control module according to the time stamp, and performs desensitization processing on sensitive data through the differential privacy algorithm. In the data storage device, establish a special database table to store sensor monitoring data and decision-making logs respectively. Add a time stamp to each data record for subsequent query and analysis.

[0099] For encrypted storage, use common encryption algorithms, such as the AES (Advanced Encryption Standard) algorithm, to encrypt the data to ensure the security of the data during storage and transmission. The differential privacy algorithm performs desensitization processing on sensitive data when the data is published or shared. For example, for some key data monitored by the sensor, under the premise of meeting a certain privacy budget, perturb the data by adding noise and other means, so that it is difficult for attackers to infer the original sensitive information from the desensitized data.

[0100] The data compression unit in the data storage module uses the wavelet transform algorithm to perform dimensionality reduction on high-frequency sampled data and achieves lossless compression through Huffman coding. When processing high-frequency sampled data, due to the large amount of data, in order to store data more efficiently, the wavelet transform algorithm is first used. The wavelet transform can decompose the data into components of different frequencies. According to the characteristics of the data, the main low-frequency components are retained, and the high-frequency components that have less impact on the overall data characteristics are removed, thereby achieving dimensionality reduction of the data.

[0101] Taking a set of time-series sensor data as an example, these data are sampled at a very high frequency. When performing wavelet transform, the original data is input into the wavelet transform function, which will perform multi-scale decomposition on the data according to the selected wavelet basis function. For example, using the Daubechies wavelet basis, after multiple layers of decomposition, the data is decomposed into an approximation component (low-frequency part) and a detail component (high-frequency part). By setting an appropriate threshold, the part of the detail component that is less than the threshold is discarded because these high-frequency details may be noise or fluctuations that have less impact on the overall trend to a certain extent. After such processing, the dimension of the data is reduced while the main characteristics of the data are retained.

[0102] After completing the dimensionality reduction by wavelet transform, Huffman coding is used to perform lossless compression on the processed data. Huffman coding is a coding method based on the frequency of data occurrence. It assigns shorter codes to data with higher occurrence frequencies and longer codes to data with lower occurrence frequencies. First, the frequency of each data value in the dimensionality-reduced data is counted, and a Huffman tree is constructed. Taking a simple data set {1, 1, 1, 2, 2, 3} as an example, the number 1 appears 3 times, the number 2 appears 2 times, and the number 3 appears 1 time. According to the construction rules of the Huffman tree, the data with lower occurrence frequencies is gradually merged, and finally a Huffman tree is constructed. In this tree, the path from the root node to the leaf node corresponding to each data value constitutes the Huffman code of that data value. For example, assuming the code of the number 1 is "0", the code of the number 2 is "10", and the code of the number 3 is "11". In this way, the data that originally needed to be stored in multiple bytes can be represented using fewer bits after Huffman coding, thereby achieving lossless compression and greatly saving the space for data storage.

[0103] It should be noted that in this text, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprising", "including" or any other variant thereof are intended to cover non-exclusive inclusion, such that a process, method, article or apparatus comprising a series of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article or apparatus.

[0104] Although the embodiments of the present invention have been shown and described, those of ordinary skill in the art can understand that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. An automatic cleaning control system for a crusher assembly line, characterized in that, It includes the following modules: The preprocessing module is used to drain the engine components, detect the residual amount of engine oil through the oil level sensor, control the oil suction device to transfer the waste engine oil to the temporary storage area, and generate a waste oil transfer instruction; The crushing control module is used to receive the completion signal from the preprocessing module, start the crusher to crush the oil-contaminated parts, adjust the rotation speed of the crushing cutter head and the feeding speed in real time based on the pressure sensor in the crushing chamber, and generate particle size distribution data of the crushed material; The dry cleaning control module is used to convey the crushed material to the dry cleaning equipment, monitor the working state of the dry cleaning chamber through the temperature sensor and the vibration sensor, and dynamically adjust the chemical agent dosing rate and the air flow circulation frequency based on the preset dry cleaning formula parameters; The logistics transmission module is used to coordinate the material transfer paths of the loader and the electromagnetic chuck, generate a material scheduling priority queue according to the stacking height of the crushed material and the capacity of the dry cleaning chamber, and optimize the transmission efficiency; The sensor monitoring module includes an oil contamination concentration detection unit, a metal debris identification unit, and an environmental temperature and humidity acquisition unit, which are used to obtain multi-dimensional data during the cleaning process in real time and generate an abnormal alarm signal; The central control module is used to integrate the data streams of the preprocessing module, the crushing control module, the dry cleaning control module, the logistics transmission module, and the sensor monitoring module, coordinate the execution timings of each module through a rule-based state machine model, and allocate system resources based on a dynamic weight scheduling algorithm.

2. The system according to claim 1, characterized in that, The crushing control module further includes: The particle size feedback unit is used to measure the particle size distribution of the crushed material in real time through a laser diffractometer, and generate a cutter head gap adjustment instruction according to a preset particle size threshold; The load balancing unit is used to dynamically allocate the load ratios of multiple groups of crushing cutter heads based on the motor current signal and the vibration spectrum analysis result to prevent local overload. Specifically, the cutter head load distribution weight is calculated by the following formula: Among them, w i is the load weight of the i-th cutter head group, I i is the amplitude of the motor current signal of the i-th cutter head group, F i is the main frequency amplitude vector of the vibration spectrum of the i-th cutter head group, and n is the cutter head group index.

3. The system according to claim 1, wherein The dry cleaning control module also includes: The chemical agent ratio optimization unit is used to calculate the mapping relationship between the chemical agent dosing amount and the oil removal rate through a non-linear regression model according to the data of the oil contamination concentration detection unit, and dynamically adjust the ratio parameters; The air flow path planning unit is used to optimize the air flow distribution by using a vortex generation algorithm based on the three-dimensional flow field simulation data in the dry cleaning chamber to improve the cleaning uniformity. The vortex generation algorithm satisfies: Among them, V s is the air flow velocity at the vortex center, α is the vortex intensity coefficient, ΔP is the pressure difference between the inside and outside of the dry cleaning chamber, and ρ is the gas density.

4. The system according to claim 1, wherein The logistics transmission module further includes: The path planning unit is used to construct the material transfer path by using a rasterized map based on the magnetic force coverage range of the electromagnetic chuck and the kinematic model of the loader, and correct the path in real time through a collision avoidance constraint algorithm; The priority dynamic adjustment unit is used to update the priority weights of the material scheduling queue by using a sliding window algorithm according to the stacking rate of the crushed material and the idle capacity of the dry cleaning chamber. The priority weight update formula is: Among them, W j is the priority weight of the j-th queue, R j is the accumulation rate of broken materials in the j-th queue, C j is the current dry cleaning bin capacity of the j-th queue, C max is the maximum capacity of the dry cleaning bin, and β is the normalization coefficient.

5. The system according to claim 1, wherein The oil contamination concentration detection unit in the sensor monitoring module uses infrared spectroscopy analysis technology to calculate the residual amount of oil contamination through the absorption rate of characteristic wavelengths, and combines the Kalman filter algorithm to eliminate environmental noise interference.

6. The system according to claim 1, wherein The dynamic weight scheduling algorithm in the central control module includes: The resource occupancy evaluation unit is used to count the real-time computing load and communication delay of each module and generate a resource occupancy matrix; A weight allocation unit, which is used to determine the priority coefficients of each module based on the matrix singular value decomposition result and optimize the weight allocation strategy by the gradient descent method.

7. The system according to claim 1, wherein It further includes: A waste recycling module, which is used to separate the metal and non-metal components in the crushed material by a magnetic separator, adjust the magnetic separation intensity based on the data of the metal debris identification unit, and optimize the separation efficiency through closed-loop feedback control.

8. The system according to claim 7, wherein The waste recycling module further includes: A hysteresis compensation unit, which is used to predict the hysteresis effect by using a polynomial fitting model according to the temperature change curve and historical magnetic flux data of the magnetic separator, and dynamically correct the magnetic separation parameters. The magnetic flux prediction model is: Φ(T) = a0 + a1T + a2T 2 where, Φ(T) is the predicted magnetic flux at temperature T, and a0, a1, a2 are fitting coefficients.

9. The system according to claim 1, characterized in that, It further includes: A data storage module, which is used to encrypt and store the original data of the sensor monitoring module and the decision-making logs of the central control module according to the time stamp, and desensitize the sensitive data by the differential privacy algorithm.

10. The system according to claim 9, wherein, The data storage module further includes: A data compression unit, which is used to perform dimensionality reduction processing on the high-frequency sampling data by using the wavelet transform algorithm and achieve lossless compression through Huffman coding.

Images