High-precision pressurized dewatering device and method for ore pulp

Abstract

The application discloses a kind of high-precision pressurized dewatering devices and methods for ore pulp, belong to ore pulp processing technical field, solve the precision of existing equipment ore pulp pressure dewatering characterization equipment is lower, and the dewatering effect of fine particle ore pulp is not ideal, high-precision pressurized dewatering device for ore pulp includes device main body, dewatering characterization integrated equipment, in-situ drying equipment, real-time monitoring module and dewatering control module;In the application, the ultra-high pressure dewatering of ore pulp is realized by the cooperation of dewatering characterization integrated equipment and dewatering control module, and the dewatering control module ensures the high precision of ore pulp dewatering mechanism, also reduces the error of fine particle dewatering characterization data, and the double-shaft transmission design of ore pulp dewatering machine breaks the stress limitation of traditional single-shaft pressurization, two dewatering transmission shafts are symmetrically distributed, can evenly disperse pressure to each area of pressure dewatering operation container, avoids the problems such as equipment deformation caused by excessive unilateral stress of traditional equipment.

Description

Technical Field

[0001] This invention belongs to the field of mineral slurry treatment technology, specifically relating to a high-precision pressure dewatering device and method for mineral slurry. Background Technology

[0002] In recent years, with the increasing mechanization of mining equipment, the particle size of raw ore entering coal washing plants has become finer, and the separation of slurry is increasingly focused on the separation of fine particles. In the coal washing industry, coal washing is a crucial step in improving coal quality, but it inevitably generates a large amount of slurry water. The efficiency and effectiveness of slurry dewatering directly affect the smooth operation of subsequent production processes and the resource recovery rate. However, traditional pressure filtration (0.6-0.8 MPa) remains the mainstream method for slurry dewatering, resulting in high product moisture content, especially for fine coal slime water, where dewatering effects are generally poor. Plate and frame filter presses are currently the most widely used pressure dewatering device for coal slime slurry. However, as the most widely used pressure dewatering device for coal slime, plate and frame filter presses have the following technical limitations:

[0003] Insufficient monitoring accuracy: It is difficult to achieve accurate real-time monitoring of the dewatering process data, and it is impossible to obtain high-precision data of key parameters such as real-time stress, cake thickness, and dewatering speed;

[0004] Pressure limitations: Although it has the stability and accuracy of high-rigidity mechanical transmission and can achieve ultra-high pressure, the pressure range of existing equipment is small, and the dewatering effect on fine particle slurry is not ideal.

[0005] The drying process is missing: the filter cake still has a high moisture content after pressure dehydration, and there is a lack of a final drying technology solution that does not damage the original state.

[0006] Therefore, in view of the problems that existing slurry pressure dewatering characterization equipment has low precision and large errors in the characterization data of fine particle dewatering; and that the pressure range of the slurry pressure dewatering process is small and the dewatering effect of fine particle slurry is not ideal, we propose a high-precision pressure dewatering device and method for slurry. Summary of the Invention

[0007] The purpose of this invention is to address the shortcomings of existing technologies by providing a high-precision pressure dewatering device and method for slurry. This invention solves the problems of low precision in existing slurry pressure dewatering characterization equipment, resulting in large errors in the characterization data for dewatering fine particles; and the small pressure range in the slurry pressure dewatering process, leading to unsatisfactory dewatering effects for fine particle slurries.

[0008] This invention is achieved by providing a high-precision pressure dewatering device for mineral slurry, the device comprising:

[0009] The device body includes a device base and an outer frame, with the outer frame fixedly mounted on the device base.

[0010] An integrated dewatering and characterization device is installed inside the base of the equipment and is used for pressurized dewatering of slurry;

[0011] The in-situ drying equipment is installed inside the outer frame of the structure and is used to dry the filter cake after pressure dehydration.

[0012] The real-time monitoring module is used to collect dehydration-related data in real time during the dehydration and drying process, preprocess the dehydration-related data, and output a real-time status set.

[0013] The dehydration control module is used to acquire a real-time status set and perform high-precision motion control on the integrated dehydration characterization equipment based on the real-time status set.

[0014] Preferably, the integrated dehydration characterization device includes:

[0015] The feeding mechanism is set on the device base and includes a raw material mixing tank, a first gas compression pump and a raw material feeding pipe. The raw material mixing tank is connected to the pressure dehydration container through the raw material feeding pipe, and the pressure dehydration container is also connected to the first gas compression pump.

[0016] The pressure dewatering operation vessel is set inside the outer frame of the structure and is used to assist in the pressure dewatering treatment of slurry;

[0017] The slurry dewatering mechanism is located inside the outer frame of the structure. The slurry dewatering mechanism is connected to the pressure dewatering operation container and is used to pressurize and dewater the slurry in the pressure dewatering operation container.

[0018] Preferably, the slurry dewatering mechanism includes:

[0019] A servo motor, which is fixedly mounted on the device base;

[0020] A servo driver is fixedly connected to the output shaft of a servo motor, and the servo driver is electrically connected to the dehydration control module;

[0021] The worm gear reducer is installed inside the device base and is connected to the servo motor;

[0022] The dehydration drive shaft is rotatably mounted inside the outer frame of the structure, and is rotatably connected to the worm gear reducer via a transmission belt.

[0023] The dewatering push assembly is slidably installed inside the outer frame of the structure, and the dewatering push part is connected to the dewatering drive shaft. The dewatering push assembly is used to pressurize and dewater the slurry in the pressure dewatering operation container and avoid surface adsorption of slag.

[0024] Preferably, the dehydration driving component includes:

[0025] A pusher sleeve is threaded onto the outer wall of the dehydration drive shaft, and the pusher sleeve is slidably connected to the outer frame of the structure.

[0026] The dehydration push rod is fixedly connected to the push sleeve;

[0027] An anti-accumulation pushing part is fixedly connected to the dewatering push rod and is used for pressurizing and dewatering the slurry;

[0028] The anti-accumulation pushing part includes:

[0029] A slurry pusher sleeve is detachably installed at the end of the dewatering push rod;

[0030] A filter cake forming seat is slidably installed in the slurry pushing sleeve. A turbulence guiding groove is provided in the filter cake forming seat. The turbulence guiding groove is used to turbulently blow off the deposits on the surface of the filter cake forming seat to prevent slag from adsorbing on the surface of the filter cake forming seat.

[0031] An auxiliary turbulence channel is formed on the side wall of the filter cake forming seat and is used to assist the operation of the turbulence guiding channel.

[0032] A reset buffer spring is fixedly connected to the filter cake forming seat. The reset buffer spring is fixedly embedded in the slurry pushing sleeve, and one end of the reset buffer spring is fixedly connected to the filter cake forming seat.

[0033] An anti-accumulation protection part is installed inside the slurry pushing sleeve. The anti-accumulation protection part is connected to the filter cake forming seat and is used to assist the filter cake forming seat in resetting and to assist in removing slag from the surface of the filter cake forming seat.

[0034] Preferably, the anti-accumulation protection part includes:

[0035] The main extrusion seat is slidably disposed within the slurry pushing sleeve, and one end of the main extrusion seat is fixedly connected to the filter cake forming seat;

[0036] The driven extrusion seat is rotatably installed inside the protective positioning seat, which is fixedly installed in the slurry pushing sleeve. The driven extrusion seat and the main extrusion seat are slidably connected.

[0037] A torsion spring support is fixedly connected to the driven extrusion seat. The torsion spring support is rotatably connected to the slurry push sleeve. A reset torsion spring is embedded in the torsion spring support. One end of the reset torsion spring is fixedly connected to the protection positioning seat.

[0038] The spring reset rod is fixedly installed inside the protective positioning seat, and one end of the spring reset rod is fixedly connected to the main compression seat.

[0039] Preferably, the in-situ drying equipment includes:

[0040] A nitrogen cylinder is fixedly installed on the device base, and the outlet of the nitrogen cylinder is connected to the second gas compression pump.

[0041] A nitrogen heating chamber, wherein the nitrogen heating chamber is used to heat nitrogen, and the nitrogen heating chamber is connected to a second gas compression pump;

[0042] The nitrogen transmission pipe is used to guide and transmit heated nitrogen gas. One end of the nitrogen transmission pipe is connected to the nitrogen heating box, and the other end is connected to the slurry pushing sleeve.

[0043] The microwave drying unit is installed inside the pressure dehydration container and is used to perform auxiliary microwave drying on the filter cake after pressure dehydration and shaping.

[0044] Preferably, the pressure dehydration container includes:

[0045] A dehydration work stand is connected to a second gas compression pump, and the microwave drying component is disposed inside the dehydration work stand.

[0046] A push guide seat is fixedly installed on both sides of the dewatering work seat. The push guide seat is slidably connected to the slurry push sleeve and is used to guide and limit the slurry push sleeve.

[0047] The microwave drying assembly includes:

[0048] An adjustable motor is fixedly installed inside the dehydration work seat;

[0049] A first gear is fixedly connected to the output shaft of the regulating motor, and the first gear is rotatably installed inside the dehydration work seat;

[0050] The second gear, which is sleeved in the dehydration work seat, rotates and meshes with the first gear. A generator mounting seat is fixedly connected to one side of the second gear.

[0051] A microwave generator, wherein the microwave generator is detachably mounted in the generator mounting base.

[0052] Preferably, the real-time monitoring module includes:

[0053] A multi-dimensional sensor group is used to collect dehydration-related data in real time during the dehydration and drying process. The multi-dimensional sensor group includes a grating displacement sensor, a stress sensor, a vibration monitoring sensor, a laser ranging displacement sensor, a temperature sensor, a liquid level sensor, a weighing sensor, and a capacitive humidity sensor.

[0054] The preprocessing unit is used to acquire dehydration-related data during the dehydration and drying process, filter and reduce noise on the dehydration-related data during the dehydration and drying process, form a time-series real-time status set, and feed the real-time status set back to the dehydration control module.

[0055] The dehydration control module includes:

[0056] The extended state space model module is used to acquire the real-time state set, and construct an extended state space model based on the real-time state set and associated data physical characteristics, time-varying parameters, and external disturbances. The extended state space model is a discrete-time state equation, and outputs the state vector, disturbance estimate, and time-varying parameters based on the extended state space model.

[0057] The model prediction cross-coupling control optimization module is used to obtain the state vector, disturbance estimate, and time-varying parameters output by the extended state space model. It uses the extended state space model to perform state prediction and error prediction, optimizes the objective function and sets constraints. It uses quadratic programming to transform the objective function into a problem that can be solved using the fast projection gradient method, and obtains the optimized state prediction and error prediction.

[0058] The neural network compensation module is used to obtain the theoretical optimal control quantity calculated by the model prediction cross-coupling control optimization module based on the extended state space model. The neural network compensation model is introduced to compensate for the model error generated in the actual control. The difference between the optimal control quantity and the theoretical control quantity of the prediction cross-coupling control optimization module forms an error sample. Using the real-time state characteristics of the system as input and the error sample as a label, the neural network compensation model is trained to learn the mapping relationship between state characteristics and error quantity. The neural network compensation model predicts the error quantity according to the current state characteristics and superimposes it into the control quantity of the prediction cross-coupling control optimization module to obtain the final control quantity, and then feeds the final control quantity back to the servo drive.

[0059] Preferably, the neural network compensation model includes a time-series data input layer and a wear level embedding layer set in parallel. The time-series data input layer is used to standardize the preprocessed time-series data. The time-series data input layer is connected to a 1D-CNN local feature extraction layer. The 1D-CNN local feature extraction layer is connected to a Transformer global dependency modeling layer. The Transformer global dependency modeling layer is also connected to an average pooling layer. The wear level embedding layer is connected to a parameter normalization layer. The parameter normalization layer is connected to a fully connected embedding layer. The fully connected embedding layer and the average pooling layer are set in parallel. The fully connected embedding layer and the average pooling layer are respectively connected to a feature fusion layer. A fully connected decision layer is set between the feature fusion layer and the output layer.

[0060] On the other hand, the present invention also provides a high-precision pressure dewatering method for slurry, the high-precision pressure dewatering method for slurry comprising:

[0061] During the feeding stage, the slurry in the raw material mixing tank is stirred, and the slurry is pumped into the dewatering working seat by the first gas compression pump. When the liquid level sensor in the dewatering working seat detects that the slurry liquid level has reached the preset liquid level, the first gas compression pump stops pumping the slurry.

[0062] During the pressurization and dewatering stage, the dewatering control module sends control parameters to the servo driver. In response to the control parameters, the servo driver starts the servo motor, which drives the dewatering push component to synchronously pressurize the slurry in the dewatering work seat in a dual-axis manner. During pressurization, dewatering-related data is collected in real time by a multi-dimensional sensor group. The dewatering-related data is preprocessed to obtain a real-time state set, which is fed back to the dewatering control module. Based on the real-time state set, a discrete-time extended state space model containing state variables, input variables, and unmodeled dynamic disturbance terms is constructed. The current state vector is obtained in real time, and the system matrix and time-varying parameters are estimated. The extended state space model module generates a control sequence in the future time domain through state prediction and error prediction, and uses a quadratic programming algorithm to continuously optimize and solve for the optimal pressurization pressure and feed flow rate at the current moment to achieve multi-variable coupled control. The error compensation amount between the model prediction value and the actual slurry dewatering effect is calculated in real time through a neural network compensation model, and the compensation amount is fed back to the model prediction cross-coupled control module to correct the current control input, obtain the final control amount, and feed the final control amount back to the servo driver.

[0063] During the drying stage, after pressurization, the nitrogen cylinder, nitrogen heating box, and microwave drying assembly are started simultaneously. The filter cake is dried by the combined action of hot nitrogen and microwave heating. The temperature inside the dehydration work seat is monitored in real time by a fiber optic temperature sensor, and the dehumidification work seat is dehumidified by a vacuum pump. Drying stops when the humidity sensor in the dehydration work seat detects a humidity of ≤5%.

[0064] Compared with the prior art, the embodiments of this application have the following main advantages:

[0065] In this embodiment of the invention, an integrated dehydration characterization device and a dehydration control module are provided. The integrated dehydration characterization device and the dehydration control module work together to achieve ultra-high pressure dehydration of the slurry. The slurry dehydration mechanism in the integrated dehydration characterization device uses the mechanical rigidity of the dual shafts to transmit pressure. Furthermore, through the servo driver and servo motor, the timely control and high precision requirements of the equipment can be improved. The dehydration control module ensures the high precision of the slurry dehydration mechanism and reduces the error of the dehydration characterization data of fine particles. At the same time, the dual-shaft transmission design of the slurry dehydrator breaks the force limitation of the traditional single-shaft pressurization. The two dehydration transmission shafts are symmetrically distributed, which can evenly distribute the pressure to all areas of the pressure dehydration container, avoiding the problems of equipment deformation caused by excessive force on one side in traditional equipment.

[0066] In this embodiment of the invention, the slurry dewatering mechanism consists of a servo motor, a servo driver, a worm gear reducer, and a dewatering drive assembly. The servo motor, servo driver, worm gear reducer, and dewatering drive assembly work together to achieve the accuracy of slurry pressurization dewatering, improve the stability and controllability of the slurry pressurization process, and during the pressurization dewatering operation, a real-time monitoring module is used to monitor and collect real-time data such as stress, displacement, vibration, flow rate, and temperature. The real-time data of stress, displacement, vibration, flow rate, and temperature are fused and transmitted to provide data support for the dewatering control module to perform real-time characterization and pressurization adjustment of the slurry pressurization dewatering process.

[0067] In this embodiment of the invention, a dewatering pushing assembly is provided, which consists of a pushing sleeve, a dewatering push rod, and an anti-accumulation pushing part. The anti-accumulation pushing part can be pushed by the pushing sleeve and the dewatering push rod, thereby realizing rapid and precise squeezing of the pressure dewatering container, improving the slurry dewatering efficiency, and preventing the adsorption of slag on the surface of the anti-accumulation pushing part when the two sets of anti-accumulation pushing parts squeeze the slurry simultaneously. The entire dewatering pushing assembly works in concert, improving the overall performance of slurry pressurized dewatering.

[0068] In this embodiment of the invention, a strategy of "extended state-space model - model prediction cross-coupling control optimization - neural network compensation module" is used, employing a master-slave motor approach to achieve synchronous and high-precision pressurization for dual-axis pressurization. From the perspective of control accuracy, the cross-coupling compensation mechanism overcomes the limitations of traditional single-axis independent control. By comparing the differences in dual-axis operating parameters in real time, compensation commands can be triggered within milliseconds, effectively eliminating synchronization errors caused by factors such as mechanical transmission backlash and load fluctuations. The master-slave motors intelligently distribute the load, effectively reducing wear on mechanical transmission components. Combined with the predictive maintenance function of the neural network, potential fault risks can be warned in advance.

[0069] In this embodiment of the invention, the in-situ drying equipment consists of a microwave drying component and a hot nitrogen air drying dehumidification component. The hot nitrogen air drying dehumidification component comprises a nitrogen cylinder, a nitrogen heating chamber, a nitrogen transmission pipe, and a negative pressure pump. The inner wall of the dehydration work seat is made of a composite three-layer material, which enhances microwave reflection and transmission, thereby improving the heating effect. The working mechanism of the microwave drying component prevents overcooling, thus enhancing the drying effect. Simultaneously, a fiber optic temperature sensor is inserted at the radial and axial center of the dehydration work seat to monitor whether the filter cake overheats during the microwave drying process and to perform self-control. The hot nitrogen air drying dehumidification component uses nitrogen as the circulating gas, which increases heating stability. A vacuum pump creates negative pressure, allowing gas flow to carry away moisture and achieving nitrogen-assisted drying. Attached Figure Description

[0070] Figure 1This is a schematic diagram of the main body of the device provided by the present invention;

[0071] Figure 2 This is a schematic diagram of the slurry dewatering mechanism provided by the present invention;

[0072] Figure 3 This is a schematic diagram of the anti-accumulation pushing part provided by the present invention;

[0073] Figure 4 This is a side view of the anti-accumulation pushing part provided by the present invention;

[0074] Figure 5 yes Figure 4 Sectional view along axis AA;

[0075] Figure 6 This is a schematic diagram of the anti-accumulation protection part provided by the present invention;

[0076] Figure 7 This is a front view of the anti-accumulation protection part provided by the present invention;

[0077] Figure 8 yes Figure 7 BB-direction sectional view;

[0078] Figure 9 This is a schematic diagram of the pressure dehydration container provided by the present invention;

[0079] Figure 10 This is a side view of the pressure dehydration container provided by the present invention;

[0080] Figure 11 yes Figure 10 CC-direction sectional view;

[0081] Figure 12 This is a schematic diagram of the structure of the microwave drying assembly provided by the present invention;

[0082] Figure 13 A schematic diagram of the master / slave motion component control principle is shown in an embodiment of the present invention;

[0083] Figure 14 A schematic diagram of multi-dimensional sensor group fusion in an embodiment of the present invention is shown;

[0084] Figure 15 The diagram illustrates the high-precision master-slave follow-drive control principle of the dehydration control module based on the error compensation mechanism.

[0085] In the diagram: 1-Main body of the device, 11-Base of the device, 12-Outer frame of the structure, 2-Feeding processing mechanism, 21-Raw material mixing tank, 22-First gas compression pump, 23-Raw material feed pipe, 3-Slurry dewatering mechanism, 31-Servo motor, 32-Worm gear reducer, 33-Servo driver, 34-Conveyor belt, 35-Dewatering drive shaft, 36-Dewatering push assembly, 361-Pushing sleeve, 362-Dewatering push rod, 37-Anti-accumulation push part, 371-Slurry push sleeve, 372-Filter cake forming seat, 373-Breakflow guide groove, 374-Auxiliary turbulence groove, 375-Reset buffer spring, 38-Anti-accumulation protection 381-Main extrusion seat, 382-Driven extrusion seat, 383-Reset torsion spring, 384-Torsion spring support seat, 385-Protective positioning seat, 386-Spring reset rod, 4-Real-time monitoring module, 5-Dehydration control module, 6-Pressure dehydration container, 61-Dehydration seat, 62-Push guide seat, 63-Microwave generator, 64-Injection port, 65-Microwave drying assembly, 651-Adjusting motor, 652-First gear, 653-Second gear, 654-Generator mounting seat, 7-In-situ drying equipment, 71-Nitrogen cylinder, 72-Second gas compression pump, 73-Nitrogen heating box, 74-Nitrogen transmission pipe. Detailed Implementation

[0086] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs; the terminology used herein in the specification of the application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application; the terms "comprising" and "having," and any variations thereof, in the specification, claims, and foregoing drawings of this application are intended to cover non-exclusive inclusion. The terms "first," "second," etc., in the specification, claims, or foregoing drawings of this application are used to distinguish different objects, not to describe a particular order.

[0087] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0088] To address the issues of low precision in existing slurry pressure dewatering characterization equipment, resulting in significant errors in data for dewatering fine particles, and a limited pressure range in the slurry pressure dewatering process leading to unsatisfactory dewatering effects on fine particle slurries, we propose a high-precision pressure dewatering device and method for slurries. In short, the device comprises a main body 1, an integrated dewatering and characterization device, an in-situ drying device 7, a real-time monitoring module 4, and a dewatering control module 5. The integrated dewatering and characterization device is housed within the device base 11 and is used for pressurized dewatering of the slurry. In this embodiment, the integrated dewatering and characterization device and the dewatering control module 5 are included. The water control module 5 works in concert to achieve ultra-high pressure dewatering of the slurry. The slurry dewatering mechanism 3 in the integrated dewatering characterization equipment uses the mechanical rigidity of the dual shafts to transmit pressure. Furthermore, the servo driver 33 and servo motor 31 can improve the timely control and high precision of the equipment. The dewatering control module 5 ensures the high precision of the slurry dewatering mechanism 3 and reduces the error of the dewatering characterization data of fine particles. At the same time, the dual-shaft transmission design of the slurry dewatering machine breaks the force limitation of the traditional single-shaft pressurization. The two dewatering transmission shafts 35 are symmetrically distributed, which can evenly distribute the pressure to all areas of the pressure dewatering container 6, avoiding the problems of equipment deformation caused by excessive force on one side in traditional equipment.

[0089] This invention provides a high-precision pressure dewatering device for mineral slurry, such as... Figure 1 As shown, the high-precision pressure dewatering device for slurry includes:

[0090] The device body 1 includes a device base 11 and an outer frame 12. The outer frame 12 is fixedly installed on the device base 11. The device base 11 can be a rectangular base or a round base structure. The outer frame 12 can be a rectangular base or a round base structure with a hollow interior. The outer frame 12 is fixedly installed on the device base 11 by welding or riveting.

[0091] An integrated dewatering and characterization device is installed inside the device base 11 and is used for pressurized dewatering of slurry;

[0092] The in-situ drying device 7 is installed inside the outer frame 12 of the structure and is used to dry the filter cake after pressure dehydration.

[0093] Real-time monitoring module 4 is used to collect dehydration-related data in real time during the dehydration and drying process, preprocess the dehydration-related data, and output a real-time status set.

[0094] The dehydration control module 5 is used to acquire a real-time state set and perform high-precision motion control on the integrated dehydration characterization equipment based on the real-time state set. The dehydration control module 5 can be an industrial computer, and the industrial computer can adopt a controller architecture using FPGA, ARM, and dedicated NPU chips, simultaneously meeting the requirements of high-precision motion control, multi-sensor data fusion, and neural network inference. The main control unit of the dehydration control module 5 can use a core component integrating FPGA and ARM to achieve parallel computing for dual-axis motion control (displacement synchronization error correction, extended state space model iteration), multi-sensor data fusion, and microwave / hot nitrogen air drying control. The edge computing unit of the dehydration control module 5 can use an NPU chip component to meet the inference requirements of 1D-CNN and Transformer. The multi-dimensional sensor group of the real-time monitoring module 4 can use an AD high-speed acquisition module. The real-time monitoring module 4 also includes a communication module integrating an EtherCAT master station, RS485, and Ethernet.

[0095] In this embodiment of the invention, an integrated dewatering characterization device and a dewatering control module 5 are provided. The integrated dewatering characterization device and the dewatering control module 5 work together to achieve ultra-high pressure dewatering of the slurry. The slurry dewatering mechanism 3 in the integrated dewatering characterization device utilizes the mechanical rigidity of a dual-axis transmission for pressurization. Furthermore, the servo driver 33 and servo motor 31 enhance the timely control and high precision of the equipment. The dewatering control module 5 ensures the high precision of the slurry dewatering mechanism 3 and reduces errors in the dewatering characterization data of fine particles. Simultaneously, the dual-axis transmission design of the slurry dewatering machine breaks through the force limitations of traditional single-axis pressurization. The two dewatering transmission shafts 35 are symmetrically distributed, which can evenly distribute the pressure to all areas of the pressure dewatering container 6, avoiding problems such as equipment deformation due to excessive force on one side in traditional equipment. Logically, the dual-axis transmission design of this invention breaks through the limitations of traditional single-axis pressurization. Through the symmetrically distributed two dewatering transmission shafts 35, the pressure is evenly distributed to all areas of the pressure dewatering container, avoiding equipment deformation and fatigue damage caused by excessive force on one side. This uniformity not only improves the overall stability of the equipment but also provides a solid foundation for the subsequent processing of fine-particle slurry. Compared with existing slurry concentration and dewatering processes (such as methods using filter cake), this invention eliminates the need for additional production steps, logically simplifies the operation process, and reduces potential points of failure.

[0096] In a further preferred embodiment of the present invention, such as Figure 1 As shown, the integrated dehydration characterization device includes:

[0097] The feeding processing mechanism 2 is mounted on the device base 11. The feeding processing mechanism 2 includes a raw material mixing tank 21, a first gas compression pump 22, and a raw material feed pipe 23. The raw material mixing tank 21 is connected to the pressure dehydration container 6 through the raw material feed pipe 23. The pressure dehydration container 6 is also connected to the first gas compression pump 22. The first gas compression pump 22 is electrically connected to the real-time monitoring module 4. The first gas compression pump 22 is fixedly installed on the device base 11 by clamps or fastening bolts. The raw material mixing tank 21 can be equipped with a stirring mechanism to fully stir the slurry, thereby ensuring that the slurry in the raw material feed pipe 23 is uniformly pumped into the pressure dehydration container 6.

[0098] The pressure dewatering operation container 6 is set inside the outer frame 12 of the structure and is used to assist in the pressure dewatering treatment of slurry;

[0099] The slurry dewatering mechanism 3 is located inside the outer frame 12 of the structure. The slurry dewatering mechanism 3 is connected to the pressure dewatering operation container 6. The slurry dewatering mechanism 3 is used to pressurize and dewater the slurry in the pressure dewatering operation container 6.

[0100] In a further preferred embodiment of the present invention, such as Figures 1-2 As shown, the slurry dewatering mechanism 3 includes:

[0101] At least one set of servo motors 31 are fixedly mounted on the device base 11. The servo motors 31 are symmetrically arranged on the device base 11, and there are two sets. The servo motors 31 are electrically connected to the dehydration control module 5.

[0102] The servo driver 33 is fixedly connected to the output shaft of the servo motor 31. The servo driver 33 is electrically connected to the dehydration control module 5. The servo motor 31 and the servo driver 33 are selected to have torque feedforward compensation function. The output current range is 0-20A and it supports real-time current adjustment in torque mode. The servo motor 31 has a rated torque of 1.5N·m and a rated speed of 3000rpm.

[0103] The worm gear reducer 32 is installed inside the device base 11 and is connected to the servo motor 31.

[0104] At least one set of dewatering drive shafts 35 are rotatably mounted inside the outer frame 12 of the structure, and the dewatering drive shafts 35 are rotatably connected to the worm gear reducer 32 via the transmission belt 34.

[0105] The dewatering push assembly 36 is slidably installed inside the outer frame 12 of the structure, and the dewatering push part is connected to the dewatering drive shaft 35. The dewatering push assembly 36 is used to pressurize and dewater the slurry in the pressure dewatering operation container 6 and avoid surface adsorption of slag.

[0106] In this embodiment of the invention, the slurry dewatering mechanism 3 consists of a servo motor 31, a servo driver 33, a worm gear reducer 32, and a dewatering drive assembly 36. The servo motor 31, servo driver 33, worm gear reducer 32, and dewatering drive assembly 36 work together to achieve the accuracy of slurry pressurization dewatering, improve the stability and controllability of the slurry pressurization process, and during the pressurization dewatering operation, the real-time monitoring module 4 monitors and collects real-time data such as stress, displacement, vibration, flow rate, and temperature, and integrates and transmits the real-time data such as stress, displacement, vibration, flow rate, and temperature, providing data support for the dewatering control module 5 to perform real-time characterization and pressurization adjustment of the slurry pressurization dewatering process.

[0107] In this embodiment, as Figures 3-5 As shown, the dehydration driving component 36 includes:

[0108] The pusher sleeve 361 is threaded onto the outer wall of the dehydration drive shaft 35, and the pusher sleeve 361 is slidably connected to the outer frame 12 of the structure. The pusher sleeve 361 can be a rectangular plate or a "T" shaped plate structure.

[0109] A dehydration push rod 362 is fixedly connected to the push sleeve 361, and the dehydration push rod 362 is fixedly connected to the push sleeve 361 by means of plugging or snapping;

[0110] The anti-accumulation pusher 37 is fixedly connected to the dewatering pusher 362 and is used to pressurize and dewater the slurry.

[0111] In this embodiment of the invention, a dewatering pushing assembly 36 is provided. The dewatering pushing assembly 36 consists of a pushing sleeve 361, a dewatering push rod 362, and an anti-accumulation pushing part 37. The anti-accumulation pushing part 37 can be pushed by the pushing sleeve 361 and the dewatering push rod 362, thereby realizing rapid and precise squeezing of the pressure dewatering container 6. This improves the dewatering efficiency of the slurry and also prevents the adsorption of slag on the surface of the anti-accumulation pushing part 37 when the two sets of anti-accumulation pushing parts 37 squeeze the slurry simultaneously. The entire dewatering pushing assembly 36 works in concert, improving the overall performance of the pressure dewatering of the slurry.

[0112] The anti-accumulation pushing part 37 includes:

[0113] The slurry pusher sleeve 371 is detachably installed at the end of the dewatering pusher 362. The slurry pusher sleeve 371 can be a hollow round seat or round sleeve structure. One end of the slurry pusher sleeve 371 is fixedly connected to the dewatering pusher 362 by plugging or snapping.

[0114] A filter cake forming seat 372 is slidably installed inside the slurry pushing sleeve 371. At least one set of turbulence guiding grooves 373 are provided inside the filter cake forming seat 372. The turbulence guiding grooves 373 are used to turbulently blow off the deposits on the surface of the filter cake forming seat 372 to prevent slag from adsorbing on the surface of the filter cake forming seat 372. The filter cake forming seat 372 can be a hollow round seat, and the surface of the filter cake forming seat 372 is polished. The turbulence guiding grooves 373 are arranged in a circumferential or matrix distribution. The turbulence guiding grooves 373 can be round grooves or square grooves.

[0115] At least one set of auxiliary turbulence channels 374 are provided. The auxiliary turbulence channels 374 are formed on the side wall of the filter cake forming seat 372. The auxiliary turbulence channels 374 are used to assist the operation of the turbulence guiding channel 373. The auxiliary turbulence channels 374 are circumferentially formed on the side wall of the filter cake forming seat 372. The setting of the auxiliary turbulence channels 374 plays a role in turbulence, which facilitates the hot air / air to blow away the residue. On the other hand, it can guide the slurry liquid entering the filter cake forming seat 372 to be discharged, thereby avoiding the residual slurry liquid in the filter cake forming seat 372.

[0116] A reset buffer spring 375 is fixedly connected to the filter cake forming seat 372. The reset buffer spring 375 is fixedly embedded in the slurry pushing sleeve 371, and one end of the reset buffer spring 375 is fixedly connected to the filter cake forming seat 372.

[0117] An anti-accumulation protection part 38 is disposed inside the slurry pushing sleeve 371. The anti-accumulation protection part 38 is connected to the filter cake forming seat 372 and is used to assist the filter cake forming seat 372 in resetting and to assist in removing slag from the surface of the filter cake forming seat 372.

[0118] In a further preferred embodiment of the present invention, such as Figures 6-8 As shown, the anti-accumulation protection part 38 includes:

[0119] The main extrusion seat 381 is slidably disposed in the slurry pushing sleeve 371, and one end of the main extrusion seat 381 is fixedly connected to the filter cake forming seat 372.

[0120] The driven extrusion seat 382 is rotatably installed in the protective positioning seat 385, which is fixedly installed in the slurry pushing sleeve 371. The driven extrusion seat 382 is slidably connected to the main extrusion seat 381. Both the main extrusion seat 381 and the driven extrusion seat 382 can be hollow arc-shaped seats or round seats, and the connection between the main extrusion seat 381 and the driven extrusion seat 382 is opened in a sliding groove.

[0121] A torsion spring support 384 is fixedly connected to the driven extrusion seat 382. The torsion spring support 384 is rotatably connected to the slurry push sleeve 371. A reset torsion spring 383 is embedded in the torsion spring support 384. One end of the reset torsion spring 383 is fixedly connected to the protective positioning seat 385. The reset torsion spring 383 ensures that the driven extrusion seat 382 and the main extrusion seat 381 quickly reset after movement.

[0122] At least one set of spring reset rods 386 are fixedly installed inside the protective positioning seat 385, and one end of the spring reset rods 386 is fixedly connected to the main pressing seat 381.

[0123] During operation, the servo motor 31 is turned on, which drives the worm gear reducer 32 to rotate. The rotation of the worm gear reducer 32 drives the dewatering drive shaft 35 to rotate. The rotation of the dewatering drive shaft 35 drives the pusher plate 361 and the dewatering push rod 362 to move. This causes the dewatering push rod 362 to drive the slurry pusher sleeve 371 and the filter cake forming seat 372 into the pressure dewatering container 6, pressurizing and dewatering the slurry inside the pressure dewatering container 6. At the same time, the filter cake forming seat 372 can compress the slurry relative to the slurry. The push sleeve 371 moves, and the filter cake forming seat 372 squeezes the main extrusion seat 381, causing the main extrusion seat 381 to drive the driven extrusion seat 382 to rotate. The driven extrusion seat 382 drives the torsion spring support seat 384 to rotate along the protective positioning seat 385. Meanwhile, the reset torsion spring 383, the spring telescopic rod, and the buffer spring react on the filter cake forming seat 372 and the main extrusion seat 381, causing the filter cake forming seat 372 and the main extrusion seat 381 to reset. This avoids residual slag on the surface of the filter cake forming seat 372 and improves the filter cake forming efficiency and dewatering efficiency.

[0124] Meanwhile, to ensure the accuracy of pressurized dewatering, in this embodiment, the servo motor 31, worm gear reducer 32, push plate 361, and servo driver 33 are symmetrically arranged in two sets about the outer frame 12, forming two sets of slurry dewatering mechanisms 3, one main and one slave. The main slurry dewatering mechanism 3 collects multi-dimensional dewatering-drying correlation data and uploads it to the dewatering control module 5. The dewatering control module 5 uses an error compensation mechanism composed of an extended state space model, model prediction cross-coupling control optimization, and neural network compensation model to perform high-precision master-slave following drive on the two sets of slurry dewatering mechanisms 3, thereby providing relatively accurate relevant process characterization data for the pressurization process of fine slurry. Figure 13 This diagram illustrates the control principle of the master / slave motion components in an embodiment of the present invention. Figure 13In this embodiment, the master / slave servo motor 31, master / slave servo driver 33, master / slave sleeve, and master / slave propulsion component can be any one of the two slurry dewatering mechanisms 3. The master / slave propulsion component is the dewatering push assembly 36 in this embodiment. The master controller is electrically connected to the dewatering control module 5. To synchronize the axial extrusion motion of the master and slave axes, a master-slave following drive scheme is proposed for the master motion part. One motor is the master station, and the other motor follows the torque or position of the master station in real time. To achieve real-time high-precision synchronization of displacement and pressure, multi-objective collaborative control of displacement synchronization (10μm level), pressure tracking (10N level), and vibration suppression is ultimately achieved. During equipment operation, in order to improve displacement and pressurization accuracy, a multi-parameter neural network compensation cross-coupling control is proposed to compensate for real-time error correction using cross-coupling. The strategy of "extended state space model - model prediction cross-coupling control optimization - neural network compensation module" is adopted, using a master-slave motor approach to achieve synchronous and high-precision pressurization for dual-axis pressurization. From the perspective of control precision advantages, the cross-coupling compensation mechanism breaks through the limitations of traditional single-axis independent control. By comparing the difference between the operating parameters of the two axes in real time, it can trigger compensation commands within milliseconds, effectively eliminating synchronization errors caused by factors such as mechanical transmission backlash and load fluctuations. The master and slave motors effectively reduce the wear of mechanical transmission components through intelligent load distribution. Combined with the predictive maintenance function of neural networks, it can provide early warning of potential fault risks and reduce complex components, further improving system stability and industrial feasibility.

[0125] In a further preferred embodiment of the present invention, such as Figure 1 As shown, the in-situ drying device 7 includes:

[0126] Nitrogen cylinder 71 is fixedly installed on the device base 11, and the outlet of nitrogen cylinder 71 is connected to the second gas compression pump 72.

[0127] A nitrogen heating chamber 73 is used to heat nitrogen gas, and the nitrogen heating chamber 73 is connected to a second gas compression pump 72.

[0128] Nitrogen transmission pipe 74 is used to guide and transmit heated nitrogen gas, and one end of nitrogen transmission pipe 74 is connected to nitrogen heating box 73, and the other end is connected to slurry pushing sleeve 371.

[0129] The microwave drying component 65 is installed inside the pressure dehydration container 6. The microwave drying component 65 is used to perform auxiliary microwave drying treatment on the filter cake after pressure dehydration and forming.

[0130] In this embodiment of the invention, the in-situ drying equipment 7 consists of a microwave drying component 65 and a hot nitrogen air drying and dehumidification component. The hot nitrogen air drying and dehumidification component comprises a nitrogen cylinder 71, a nitrogen heating chamber 73, a nitrogen transmission pipe 74, and a negative pressure pump. The inner wall of the dehydration work seat 61 is made of a composite three-layer material, which enhances microwave reflection and transmission, thereby improving the heating effect. The working mechanism of the microwave drying component 65 prevents overcooling, thus enhancing the drying effect. Simultaneously, a fiber optic temperature sensor is inserted at the radial and axial center of the dehydration work seat 61 to monitor whether the filter cake overheats during the microwave drying process and to perform self-control. The hot nitrogen air drying and dehumidification component uses nitrogen as the circulating gas, which increases heating stability. A vacuum pump creates negative pressure, allowing gas flow to carry away moisture and achieving nitrogen-assisted drying. The in-situ drying equipment 7, combining hot nitrogen transmission and the microwave drying component 65, directly and synergistically processes the filter cake after pressurized dehydration: hot nitrogen uniformly evaporates surface moisture, and microwaves penetrate internal moisture, avoiding damage to the filter cake structure. This in-situ design is logically closely integrated with the pressurization stage, shortening the transition time from dewatering to drying and reducing the pollution and energy consumption risks associated with external transportation. Compared to existing slurry drying systems (such as those using activated alumina) or seabed mining dewatering systems, this invention embeds drying into the main body of the device, forming an integrated process. This logically optimizes resource utilization and improves the final recovery rate of fine-particle slurry. Overall, with a dual-shaft drive providing the mechanical foundation and monitoring and control ensuring accuracy, the in-situ drying completes a closed loop. This achieves full-chain optimization of slurry treatment, promotes equipment reliability and industrialization potential, avoids the risks of unnecessarily complex solutions, and ensures the solution is more practical in real-world applications.

[0131] In a further preferred embodiment of the present invention, such as Figures 9-11 As shown, the pressure dehydration container 6 includes:

[0132] A dehydration work stand 61 is connected to a second gas compression pump 72. A microwave drying assembly 65 is disposed inside the dehydration work stand 61. The outer structure of the dehydration work stand 61 is a stainless steel cylindrical container to provide mechanical structural strength during pressurization and microwave shielding during drying. The main structure of the dehydration work stand 61 utilizes a composite inner layer to cover the entire inner wall, including the inner surface of the column and the inner surface of the pressurization end cap, to prevent the coal slime filter cake from directly contacting the metal wall. The outer layer of the composite outer layer is made of alumina ceramic, the inner layer is made of polytetrafluoroethylene (PTFE), which is corrosion-resistant and easy to clean. The middle layer uses a double-layer aluminum reflective ring to enhance the heating effect. The double-layer aluminum reflective ring of the middle layer is based on the arrangement of the microwave drying assembly 65 described below. A material inlet 64 is opened on the side wall of the dehydration work stand 61, which is connected to the raw material inlet pipe 23.

[0133] At least one set of push guide seats 62 are fixedly installed on both sides of the dewatering operation seat 61. The push guide seats 62 are slidably connected to the slurry push sleeve 371 and are used to guide and limit the slurry push sleeve 371.

[0134] Among them, such as Figure 12 As shown, the microwave drying assembly 65 includes:

[0135] An adjusting motor 651 is fixedly installed inside the dehydration work seat 61;

[0136] A first gear 652 is fixedly connected to the output shaft of the regulating motor 651, and the first gear 652 is rotatably installed in the dehydration work seat 61;

[0137] The second gear 653 is rotatably sleeved in the dehydration work seat 61. The second gear 653 meshes with the first gear 652 for transmission. A generator mounting seat 654 is fixedly connected to one side of the second gear 653.

[0138] At least one set of microwave generators 63 are provided, which are detachably installed in the generator mounting base 654. The magnetron microwave emitting unit of the microwave drying microwave generator 63 is embedded and mounted against the inner side wall of the dehydration work base 61. Two sets of three magnetron microwave emitting units (set 1 and set 2) can be designed. The positions of the three magnetron microwave emitting units in each set are at 1 / 3 and 2 / 3 of the axial direction, respectively. The microwave power of the microwave generator 63 is adjustable from 0 to 1000W, and the operating frequency range is 2450MHz ± 50MHz. The magnetrons of set 1 and set 2 are axially offset by 60° from the magnetrons of set 2 and set 1, respectively. The working mechanism of the magnetron microwave emitting unit is that set 1 operates from 0 to 10s, set 2 operates from 10 to 20s (to reduce interference), and the magnetron emitting units in the same set are activated in turn in sequence. When the temperature is > 80℃, the microwave power is reduced to 600W. To prevent filter cake or liquid water from contaminating the magnetron microwave emitting unit during the drying process, a superhydrophobic nano-coating can be applied to the wavefront surface. To fit the overall design of the dehydration device, the magnetron and waveguide are integrated and embedded in an opening in the side wall of the generator mounting base 654. A miniature microwave water-cooling device is also integrated behind the magnetron to prevent overheating. A fiber optic temperature sensor is inserted at the radial and axial center of the container to monitor whether the filter cake overheats during the microwave drying process; the data is transmitted to the dehydration control module 5 for automatic power reduction or shutdown. A humidity sensor is placed in the vacuum pump pipeline to monitor the progress of filter cake drying, and the dehydration control module 5 provides automatic control. Metal mesh gaskets are added to the seams of the generator mounting base 654, and a metal filter is installed at the end cap to prevent microwave leakage.

[0139] In this embodiment, to address the requirements of in-situ drying in a closed cylindrical dewatering work station 61 (the coal slime filter cake is axially compressed and tangentially positioned in the middle of the container, and cannot be moved), and to ensure as uniform a drying process as possible while preventing overheating of the slurry filter cake from affecting its properties, a hot nitrogen air drying and dehumidification assembly consisting of a microwave drying component 65, a nitrogen cylinder 71, a nitrogen heating chamber 73, a nitrogen transmission pipe 74, and a negative pressure pump is provided. The specific microwave drying process is as follows: Preheating and Start-up Stage: The filter cake drying temperature is slowly increased, and the microwave generator 63 is alternately started and operated. The operating logic is as follows: the microwave generator 63 has a power of 600W, the nitrogen temperature of the hot nitrogen air drying and dehumidification assembly is 90℃, and the flow rate is 0.3m³ / h. 3 / h, vacuum degree -0.05MPa, the next stage begins when the humidity sensor detects that the exhaust humidity has increased from 12%RH to 28%RH.

[0140] Rapid drying stage: Power is increased to 800W, nitrogen air drying module, nitrogen temperature 110℃, flow rate 0.6m³ / h, vacuum degree -0.07MPa. When the humidity monitoring drops sharply from the peak to 18%RH, it enters the next stage.

[0141] Final fine drying: Reduce microwave generator power from 63W to 500W, use nitrogen drying module, nitrogen temperature 110℃, flow rate 0.4m³ / h, vacuum degree -0.06MPa. When humidity ≤5%RH for 30 consecutive seconds, trigger shutdown and allow to stand for 2 minutes.

[0142] In this embodiment of the invention, the hot air drying and dehumidification scheme of the hot nitrogen air drying and dehumidification assembly includes two aspects: one is supplying hot air, and the other is creating negative pressure. During operation, the second gas compression pump 72 compresses nitrogen gas and delivers it along the nitrogen gas transmission pipe 74 to the pressure dehydration container 6. The gas flows through the nitrogen heating box 73 and is heated. During the drying process, hot nitrogen gas flows into the pressure dehydration container 6. On the other side of the end cap inside the pressure dehydration container 6, a vacuum pump extracts a mixture of nitrogen gas and water vapor, creating negative pressure. To adapt to the high-precision pressurized dewatering device for this slurry, the gas transmission pipeline is not exposed externally, but part of it is also embedded in the dewatering drive component 36; a humidity sensor is installed in the pipeline 30-50cm before the vacuum pump inlet, and the humidity sensor is linked with the dewatering control module 5, the nitrogen heating box 73, and the vacuum pump. When the humidity is ≤5%RH for 30s, the dewatering control module 5 turns off the microwave generator 63, the hot nitrogen temperature drops to 80℃, and continues to air dry for 5 minutes (to remove residual water vapor), after which the drying ends.

[0143] In a further preferred embodiment of the present invention, the real-time monitoring module 4 includes:

[0144] A multi-dimensional sensor group is used to collect dehydration-related data in real time during the dehydration and drying process. The multi-dimensional sensor group includes, but is not limited to, grating displacement sensors, stress sensors, vibration monitoring sensors, laser ranging displacement sensors, temperature sensors, liquid level sensors, weighing sensors, and capacitive humidity sensors.

[0145] The preprocessing unit is used to acquire dehydration-related data during the dehydration and drying process, filter and reduce noise on the dehydration-related data during the dehydration and drying process, form a time-series real-time status set, and feed the real-time status set back to the dehydration control module 5.

[0146] in, Figure 14This diagram illustrates the fusion of a multi-dimensional sensor group in an embodiment of the present invention. To meet equipment requirements, the grating displacement sensor uses a repeatability accuracy ≤2μm and employs a dual-reading-head design (one set for the master axis and one for the slave axis). A signal conditioning module is added between the grating ruler and the dehydration control module 5 to convert the differential signal output by the grating ruler into a single-ended signal. Simultaneously, a 200Hz low-pass filter removes high-frequency electromagnetic interference. The support frame is made of aluminum alloy (lightweight and high rigidity), and the shock-absorbing pads are made of polyurethane (damping coefficient 5Ns / m) to reduce errors. Displacement data is transmitted to the dehydration control module in real time via an RS485 bus. 5. The grating displacement sensor is equipped with a protective cover and other protective designs to achieve an IP67 protection rating; the high-precision stress sensor is a Wheatstone full-bridge column-type tension / compression sensor with a range of 0-40000N, an accuracy of 0.02%FS, and a nonlinear error ≤0.01%FS. The sensor is embedded in the center of the extrusion surface of the filter cake forming seat 372, and a titanium alloy gasket is added to the filter cake forming seat 372. The output signal is converted by an AD converter (sampling rate 1kHz) and then connected to the dewatering control module 5. The protective cover and other protective designs achieve an IP67 protection rating; the vibration monitoring sensor is selected from... The piezoelectric accelerometer has a measurement range of 0-5g, a sensitivity of 100mV / g, a frequency response of 1-1000Hz, and a lateral sensitivity of ≤5%. It is primarily installed on the end face of the master / slave push plate 361 near the dehydration drive shaft 35, with the installation point ≤20mm from the axis of the dehydration drive shaft 35. To reduce the influence of other factors, a wavelet noise reduction module is added between the sensor and the dehydration control module 5. The vibration signal is decomposed into three levels using the db4 wavelet basis function to filter out environmental vibrations and electromagnetic interference, retaining the effective vibration frequency band of 10-500Hz. Simultaneously, the vibration sensor... The sample trigger signal is bound to the grating ruler sampling signal to ensure that the timestamps of displacement and vibration data are consistent. A laser ranging displacement sensor is used for belt tension detection, with a measurement range of 0-10mm, a resolution of 1μm, and a measurement accuracy of ≤±2μm. The sensor is fixed above the conveyor belt 34, and high-reflectivity markers (5mm diameter, reflectivity ≥80%) are affixed to the surface of the conveyor belt 34. In the aforementioned in-situ drying equipment 7, a temperature sensor is also embedded to monitor the real-time drying temperature of the pressure dehydration container 6 and prevent overload. Due to the special operating environment, a fiber optic temperature probe is selected. Similarly, a liquid level sensor is installed at the inlet of the pressure dehydration container 6, using an RF admittance liquid level switch sensor for automatic control of the feeding operation.The sensor data fusion design adopts system clock synchronization triggering, using the 1ms cycle clock of the dehydration control module 5 as a reference to synchronously trigger the sampling of the grating ruler, pressure sensor, and vibration sensor (sampling time deviation ≤10us). The laser rangefinder (500Hz) achieves synchronization with the main clock through clock frequency division (frequency division error ≤0.1ms), ensuring that the timestamps of all sensor data are consistent. A high-precision electromagnetic force compensated weighing sensor is used to monitor the real-time amount of pressure dehydration during the operation. The in-situ drying part uses a capacitive humidity sensor with a temperature range of 0℃-180℃ (due to the long-term working temperature ≤120℃), an accuracy of ±1.5%RH at 0-90%RH, and an accuracy of ±0.8%RH at 0-20%RH, which meets the requirements for working in negative pressure and has a protection level of IP65 or higher.

[0147] This invention integrates the synergistic effect of a real-time monitoring module 4 and a dewatering control module 5. The real-time monitoring module 4 collects dewatering-related data (such as displacement, stress, and vibration) through a multi-dimensional sensor array and preprocesses it to form a time-series state set, providing accurate input to the control module. The dewatering control module 5 captures system dynamics (such as time-varying parameters and external disturbances) based on an extended state-space model, combines model prediction with cross-coupling control to optimize multivariables (such as dual-axis synchronous pressurization), and uses neural network compensation to correct errors in real time. This multi-layered synergy is not a simple superposition but logically complementary: the state-space model provides global prediction, cross-coupling ensures synchronization, and neural compensation fine-tunes unmodeled disturbances, thereby achieving high-precision motion control. Compared to existing high-pressure slurry ablation or suspension dewatering methods, the control framework of this invention non-obviously integrates mechanical stiffness and intelligent algorithms, logically reducing errors in fine particle dewatering data, improving process controllability and consistency, and avoiding the inefficiency of traditional equipment relying on manual adjustments.

[0148] In a further preferred embodiment of the present invention, such as Figure 15 The diagram illustrates the high-precision master-slave follow-drive control principle of the dehydration control module 5 based on an error compensation mechanism. The dehydration control module 5 includes:

[0149] The extended state space model module is used to acquire the real-time state set, and construct an extended state space model based on the real-time state set and associated data physical characteristics, time-varying parameters, and external disturbances. The extended state space model is a discrete-time state equation, and outputs the state vector, disturbance estimate, and time-varying parameters based on the extended state space model.

[0150] The real-time state set includes multi-faceted and multi-dimensional data, incorporating the physical characteristics, time-varying parameters, and external disturbances of the mechanical system into the state-space model (multi-physical parameters). The general form of the extended state-space model is a discrete-time state equation (sampling period of 1ms) as follows:

[0151] ,

[0152] The state vector x(k) contains nine core dynamic variables, including real-time two-axis displacement data x1 and x2 (grating ruler displacement sensor), real-time two-axis motion velocity data v1 and v2 (displacement signal differential smoothed by Kalman filtering), real-time two-axis sleeve extrusion pressure data p (high-precision stress sensor), vibration acceleration data a1 and a2 of the sleeve propulsion shaft (piezoelectric acceleration sensor), real-time conveyor belt tension s (laser ranging displacement sensor), and ball screw wear coefficient ω (calculated based on vibration spectrum characteristics, 0-1 new part - severely worn).

[0153] ,

[0154] The input vector u(k) represents the control inputs of the two servo motors 31, determining the system's motion state. u1 and u2 are the torque commands (in N·m) of the master and slave servo motors 31, which are converted into current outputs by the servo driver 33 to control the motor's output torque.

[0155] ,

[0156] Time-varying parameters Including the wear of the ball screw (Value range 0-1, based on the energy of ball screw vibration in the 100-500Hz frequency band) (k is the calibration coefficient) and belt aging coefficient (Value range 0-1, (k is the aging coefficient) and temperature influence coefficient (Value range 0-0.2) (k=0.01 / ℃).

[0157] The uncontrollable disturbance term d(k) includes three types of random disturbances, d f (Frictional disturbance, Random fluctuations in static friction and stick-slip friction at low speeds are calculated using the residuals of the Stribeck model; d v (Vibration interference, g) The effect of external environmental vibration (such as ground vibration) on the dehydration pusher 362, extracted by high-frequency components of the vibration sensor; d p (Pressure fluctuation, N) refers to the random pressure fluctuation caused by the surface roughness of the contact plate 361, calculated using the pressure signal filtering residual.

[0158] The output vector y(k) corresponds directly to the sensor data. ,

[0159] It is the state transition matrix, which represents the dynamic coupling relationship between state variables (such as the effect of velocity on displacement, the effect of vibration on pressure), and its elements follow... change; It is an input matrix, representing the influence of control variables on state variables (such as the influence of torque commands on speed), and its elements follow... change; The output matrix is ​​obtained by selecting x1, x2, and p from the state vector. The relevant physical parameters mentioned above are then obtained. and The specific form is as follows:

[0160] ,

[0161] ,

[0162] Where, m 1\2 It is to promote the equivalent quality of the 361-degree cladding. It is the damping coefficient that varies with temperature. Increase (increase) , For vibration-pressure coupling coefficient, For pressure damping coefficient, For equivalent stiffness, Vibration attenuation coefficient, The tensile attenuation coefficient increases with belt aging. For wear rate coefficient, Transmission efficiency coefficient The coupling coefficient varies with belt aging.

[0163] in It is a nonlinear perturbation function, representing the combined effects of nonlinear factors such as friction, vibration, and wear. The core component is frictional nonlinearity; at low speeds (v < 5 mm / s), the frictional force exhibits stick-slip characteristics. The nonlinear perturbation function is modeled as follows: ,in The velocity threshold (approximately 0.5 mm / s), d f The friction disturbance amplitude; wear-vibration coupling, when the wear of the dehydration drive shaft 35 intensifies, the vibration energy is significantly enhanced in a specific frequency band (such as the meshing frequency of 200Hz), and the nonlinear disturbance function is modeled as follows: , where f m The meshing frequency of the dehydration drive shaft is 35, d v The vibration disturbance amplitude is shown; the pressure-displacement nonlinearity means that the relationship between pressure and displacement is nonlinear when the plates are in contact. The nonlinear disturbance function is modeled as follows: .

[0164] The model prediction cross-coupling control optimization module is used to obtain the state vector, disturbance estimate, and time-varying parameters output by the extended state space model. It uses the extended state space model to perform state prediction and error prediction, optimizes the objective function and sets constraints. It uses quadratic programming to transform the objective function into a problem that can be solved using the fast projection gradient method, and obtains the optimized state prediction and error prediction.

[0165] During the "rolling optimization" process of the model prediction cross-coupling control optimization module, the extended state-space model outputs three key pieces of information to the module: the state vector. The real-time states of the main motion kit, including displacement, velocity, pressure, vibration, belt tension, and wear of the dehydration push rod 362, are used as the initial state parameters for the model prediction cross-coupled control optimization module; time-varying parameters This includes quantifying the effects of equipment wear, aging, and temperature, thereby correcting the system matrix of the predictive cross-coupled control optimization module. Perturbation estimation This includes real-time estimation of frictional disturbances, vibration disturbances, and pressure fluctuations, and the "feedforward compensation term" of the model predicts the cross-coupled control optimization module to offset disturbances in advance in the prediction model.

[0166] Based on the extended state-space model, the state is predicted for the next N=15 steps, and the corresponding synchronization error prediction is calculated. Multi-dimensional cross-coupling error is key to the predictive cross-coupling control optimization module. To integrate the displacement, pressure, and vibration of this device, the error vector is set as follows: ,

[0167] ,

[0168] in This refers to the displacement synchronization error (including velocity difference). Take 20, Choosing 5 ensures that the position and speed of the axial movement of the master / slave pusher plate 361 are synchronized; For pressure tracking error, Set the value to 0.2 to ensure that the extrusion pressure accurately tracks the command value; To suppress vibration error, A value of 0.1 is used to suppress propeller shaft vibration;

[0169] The model equations for the predicted cross-coupling control optimization module for N steps are as follows (N is typically taken as 15):

[0170] ,

[0171] Where i = 1, 2, 3, ... N;

[0172] Based on the state prediction results, the predicted synchronization error for the next 15 steps is calculated, and the following results are obtained:

[0173] ,

[0174] The objective function is optimized and constraints are set. Then, quadratic programming (QP) is used to transform it into a problem that can be solved using the fast projected gradient method.

[0175] The objective function formula for predicting the cross-coupling control optimization module is as follows:

[0176] + ;

[0177] The priority order is: displacement synchronization (100) > pressure tracking (50) > vibration suppression (20). ;

[0178] For smooth control and to prevent sudden changes in control values Preventing excessive torque commands The P term in the last term of the above function is the terminal weight matrix, which is solved using the linear matrix inequality (LMI) to ensure the stability of the optimization.

[0179] The constraints are physical safety boundaries. The upper limit of the motor output torque (to avoid overcurrent) and the control variable amplitude constraints are... ,( ),in , To avoid mechanical shock caused by sudden torque changes, the rate of change of the control variable is constrained as follows: ,( ),in( ); safe range of displacement and pressure,

[0180] Its output constraints are:

[0181] ,

[0182] From the above, we obtain the objective function and constraints, and then solve them:

[0183] Such that AcU≤bc, where Let H be the control sequence to be optimized, and g be the coefficient matrices of the QP problem. c b c The constraint matrix is ​​then used; the fast projection gradient method is then employed to quickly solve for the optimal control sequence U. * .

[0184] The above obtains the optimal sequence. The first step of the optimal sequence is executed, the remaining sequence is discarded, and the process of obtaining the state and parameters, predicting the state and error, optimizing the state and error prediction, and rolling optimization is repeated in the next cycle (k+1).

[0185] The neural network compensation module is used to obtain the theoretical optimal control quantity calculated by the model prediction cross-coupling control optimization module based on the extended state space model. The actual optimal control quantity is the control quantity corresponding to the minimum synchronization error that can control the dual-axis displacement synchronization error within 10μm and the pressure synchronization error within 10N under the current operating conditions. It is obtained by combining the system operating conditions (dual-axis relative compression, multi-sensor data feedback) during the system debugging phase. The neural network compensation model is introduced to compensate for the model error generated in the actual control. The difference between the optimal control quantity and the theoretical control quantity of the prediction cross-coupling control optimization module forms an error sample. With the real-time state characteristics of the system as input and the error sample as label, the neural network compensation model is trained to learn the mapping relationship between state characteristics and error quantity. The neural network compensation model predicts the error quantity according to the current state characteristics and superimposes it into the control quantity of the prediction cross-coupling control optimization module to obtain the final control quantity, and feeds the final control quantity back to the servo drive 33.

[0186] In this embodiment of the invention, the actual optimal control quantity is obtained as follows: During the system debugging phase, an initial sample library of actual optimal control quantities is established by covering typical working conditions through an automatic optimization algorithm. Specifically, the preset key working conditions include slurry concentration (10%-40%, 5% interval), particle size (50μm-500μm, 50μm interval), ambient temperature (0℃-50℃, 10℃ interval), reducer wear coefficient (0-1, 0.1 interval), belt tension (0-5mm, 0.5mm interval), preset pressure (500N-10000N, 200N interval; 10000N-30000N, 500N interval), and movement speed (0mm / s-20mm / s, 5mm / s interval). These working conditions are adjusted using a high and low temperature test chamber, a wear reducer replacement device, and a belt tension adjustment mechanism. The equipment is started according to the operating conditions, and 10-dimensional characteristic parameters are monitored in real time. When the parameter fluctuation is ≤2% within 30 consecutive seconds, the operating condition is considered stable. The theoretical control quantity of the cross-coupling control optimization output is predicted using the above model, and the torque output of the dual shafts at this time is recorded. Data is continuously collected for a certain period of time (3s), and the average displacement difference and average pressure difference are calculated as a benchmark. The gradient descent algorithm is adopted to... For the objective function ( (for vibration error), with theoretical control quantity Optimization was performed iteratively within the rated torque range, with an iterative adjustment step size set to 0.5% of the rated torque. Data was collected for 2 seconds after each adjustment, and the objective function value was recalculated. When the average displacement difference was ≤10μm and the average pressure difference was ≤10N, the vibration amplitude was... 0.1g (accelerometer unit) is recorded as the current actual control quantity 1. The above process is repeated 3 times for the same operating condition. If the maximum deviation of the actual control quantity in the 3 times is ≤1% of the rated torque, the average value is taken as the final actual control quantity for that operating condition, ultimately generating an 8-dimensional feature mapping relationship library. For uncovered operating conditions, the K-nearest neighbor algorithm is used to match the most similar operating condition from the offline library, using the theoretical control quantity as the initial control quantity. If the two-axis displacement difference and pressure difference in each cycle satisfy >15μm and >15N for 3 consecutive cycles respectively, a particle swarm optimization algorithm is used [particle number 20, search range is the initial value]. The rated torque, the number of iterations is less than 11, and the objective function A rapid optimization process is performed. Finally, the two optimized actual control variables are fused to form the neural network training dataset.

[0187] The prediction error is influenced by both local transient errors and global time-varying errors. Based on this requirement, a 1D-CNN+Transformer hybrid neural network structure is adopted, characterized by the following components: The input data receives 10-dimensional time-series feature data consisting of two-axis displacement difference, pressure difference, slurry concentration, particle size, vibration amplitude, reducer wear coefficient, belt tension, motor temperature, reducer temperature, and ambient temperature, as well as 9-dimensional static parameter data including worm gear wear level*2, backlash value*2, belt wear amount*2, belt stiffness coefficient*2, and cumulative running time*1. The time-series feature parameters are standardized using a sliding window to form a 50-step time-series matrix. The worm gear wear level is mapped to a 4-dimensional vector through an embedding layer. The backlash value and belt wear amount undergo logarithmic transformation, and all parameters are normalized to the [0, 1] interval.

[0188] The sliding window standardization is as follows ( ) (The mean and standard deviation of the i-th feature within a 50-step window).

[0189] ,

[0190] In this embodiment, the neural network compensation model includes a time-series data input layer and a wear level embedding layer set in parallel. The time-series data input layer is used to standardize the preprocessed time-series data. The time-series data input layer is connected to a 1D-CNN local feature extraction layer. The 1D-CNN local feature extraction layer is connected to a Transformer global dependency modeling layer. The Transformer global dependency modeling layer is also connected to an average pooling layer. The wear level embedding layer is connected to a parameter normalization layer. The parameter normalization layer is connected to a fully connected embedding layer. The fully connected embedding layer and the average pooling layer are set in parallel. The fully connected embedding layer and the average pooling layer are respectively connected to a feature fusion layer. A fully connected decision layer is set between the feature fusion layer and the output layer.

[0191] It should be noted that the specific time-series data processing flow is as follows:

[0192] Input layer: used to receive 50 steps of time-series data of the above 10-dimensional features, with each step corresponding to a 1ms control period;

[0193] The 1D-CNN local feature extraction layer includes convolutional layer 1, which contains 32 convolutional kernels of varying sizes. With a stride of 1 and padding of 2, it captures vibration peaks and sudden displacement differences within 5ms, local transient features; it uses batch normalization to standardize the output of each convolutional kernel and ReLU activation to address the problem of linear models being unable to fit complex errors; max pooling layer - pooling kernel size. , used to compress data volume by preserving maximum features; Convolutional layer 2, which includes 64 convolutional kernels, size With a step size of 1 and padding of 1, it captures fine-grained features such as temperature gradient changes and small fluctuations in pressure difference within 3ms; it averages the 25 time dimensions, compresses them into 1 dimension (global average pooling), extracts global statistics of local features, and outputs a 64-dimensional local feature vector.

[0194] Transformer global dependency modeling layer: used to add temporal location information to 64-dimensional local feature vectors, using sine and cosine for location encoding.

[0195] ,

[0196] ,

[0197] The output dimension is 64, and location information has been added.

[0198] By utilizing a multi-head self-attention mechanism, the global dependencies between features are modeled, and the 64-dimensional features are divided into eight 8-dimensional subspaces. Attention weights are calculated for each subspace.

[0199] ,

[0200] By utilizing residual connections and layer normalization, we can improve training stability, prevent gradient vanishing, and accelerate model convergence.

[0201] ,

[0202] By utilizing feedforward neural networks, the nonlinear fitting capability is enhanced, and the features output by the attention mechanism are processed.

[0203] The input data is fed into the encoder module (multi-head self-attention mechanism + residual connection and layer normalization + feedforward neural network) for two layers of encoder stacking.

[0204] Output layer: Includes a fully connected layer 1, which uses ReLU activation to compress the feature dimension and extract key information. The output dimension decreases from 64 to 32. The fully connected layer 1 maps the features to two-axis compensation values, decreasing the output dimension from 32 to 2. The compensation values ​​are constrained by the Tanh activation function. Within the rated torque range of the motor: To avoid overcompensation.

[0205] Meanwhile, the specific learning and training methods for the neural network compensation model include offline mapping relationship training and online real-time compensation execution:

[0206] Offline mapping training:

[0207] The data acquisition system was built to collect data under various operating conditions, including different temperatures, different wear levels in the reducer, and belt tension, with 50 steps as input. A 10-dimensional time-series feature matrix was used. Since this device is suitable for fine-particle slurry, the sample size was 5000 groups, which were divided into training set, validation set and test set according to a ratio of 7:2:1.

[0208] Model training requires using the AdamW optimizer and setting weight decay. The initial learning rate is The loss function is: ( ) Loss guarantee and compensation accuracy (Excessive regularization suppression compensation)

[0209] ,

[0210] Training stops when the validation set loss does not decrease for 10 consecutive epochs, or reaches 50 epochs.

[0211] The model was evaluated using a test set, with key metrics including compensation error ≤0.5μm and ≤0.5N (displacement equivalent control quantity and pressure equivalent control quantity). The inference speed was deployed on the edge NPU, with a single-sample inference time ≤0.8ms.

[0212] Online real-time compensation operation:

[0213] Sensors collect data, which is processed by Kalman filtering for displacement and pressure, and by wavelet denoising for vibration. Ten-dimensional features are extracted, and a sliding window is constructed: each step updates the feature data of the most recent 50 cycles, forming a temporal matrix, which is then standardized. The temporal matrix (1D-CNN layer) is then convolved, pooled, and output to obtain local features. Position encoding (Transformer layer) is added to the local features, which are then processed by a two-layer encoder to output global features. The output layer maps the global features to compensation values, which are then constrained within a safe range using Tanh activation.

[0214] The control inputs are fused, and the final control input = predicted cross-coupling control optimization control data + neural network compensation control input. The result is output to the servo driver 33. Every 500 cycles, new error samples are calculated using the real-time synchronization error in the predicted cross-coupling control optimization module. The Elastic Weight Consolidation (EWC) method is used to update the top 30% weights of the network.

[0215] On the other hand, the present invention also provides a high-precision pressure dewatering method for slurry, the high-precision pressure dewatering method for slurry comprising:

[0216] S10, feeding stage, the slurry in the raw material mixing tank 21 is stirred, and the slurry is pumped into the dewatering work stand 61 by the first gas compression pump 22. When the liquid level sensor in the dewatering work stand 61 detects that the slurry liquid level has reached the preset liquid level, the first gas compression pump 22 stops pumping the slurry.

[0217] S20, during the pressurization and dewatering stage, the dewatering control module 5 sends control parameters to the servo driver 33. The servo driver 33 responds to the control parameters by starting the servo motor 31, which in turn drives the dewatering push assembly 36 to synchronously pressurize the slurry in the dewatering work seat 61 along two axes. During pressurization, a multi-dimensional sensor group collects dewatering-related data in real time. The dewatering-related data is preprocessed to obtain a real-time state set, which is then fed back to the dewatering control module 5. Based on the real-time state set, a discrete-time extended state space containing state variables, input variables, and unmodeled dynamic disturbance terms is constructed. The system uses a time-space model to obtain the current state vector and estimate the system matrix and time-varying parameters in real time. The extended state-space model module generates the control sequence in the future time domain through state prediction and error prediction. It uses a quadratic programming algorithm to solve the optimal pressurization pressure and feed flow rate at the current moment to achieve multivariable coupled control. The neural network compensation model calculates the error compensation amount between the model prediction value and the actual slurry dewatering effect in real time, and feeds the compensation amount back to the model prediction cross-coupled control module to correct the current control input, obtain the final control amount, and feed the final control amount back to the servo driver 33.

[0218] Specifically, in the initialization stage of the slurry dewatering mechanism 3: the servo driver 33 is powered on and preheated for 1 minute, the grating displacement sensor returns to zero, the control push plate 361 is moved to the mechanical origin, the initial reading of the grating ruler is recorded (main axis x1 = 0.000 mm, slave axis x2 = 0.001 mm), and a compensation command is sent through the FPGA to correct the slave axis displacement to 0.000 mm, with an initial synchronization error ≤ 1 μm; the model initialization parameters are set (wear coefficient of dewatering drive shaft 35, belt aging coefficient, temperature coefficient, and initial values ​​of disturbance terms, etc.).

[0219] During the pressurized start-up phase: In the low-pressure range, the motor outputs torque according to preset requirements. The motor tracks the main motor torque in real time (torque following error ≤ 0.03 N·m). The model prediction cross-coupling control module calculates the error and corrects the output torque. In the medium-pressure range, a neural network compensation model is introduced to compensate. The compensation amount is updated every 50 ms. At the same time, the prediction step number of the model prediction control is N=15.

[0220] Pressure stabilization phase: The pressure stabilizes at the set pressure (e.g., corresponding to an extrusion pressure of 35000N). The stress sensor monitors pressure fluctuations in real time. When the pressure drops to 34992N (deviation 8N), the neural network compensation model predicts and controls the output torque increment of +0.02N·m. Within 1ms, the pressure recovers to 35000N. Displacement synchronization error control: Corrected every 1ms through a closed-loop optical grating ruler.

[0221] S30, Drying stage: After pressurization, nitrogen cylinder 71, nitrogen heating box 73, and microwave drying assembly 65 are started simultaneously. The filter cake is dried by the combined action of hot nitrogen and microwave heating. The temperature inside the dehydration work seat 61 is monitored in real time by a fiber optic temperature sensor, and the dehumidification work seat 61 is dehumidified by a vacuum pump. Drying stops when the humidity sensor in the dehydration work seat 61 detects a humidity of ≤5%.

[0222] In summary, this invention provides a high-precision pressure dewatering device and method for mineral slurry. In this embodiment, an integrated dewatering characterization device and a dewatering control module 5 are provided. Through the coordinated operation of the integrated dewatering characterization device and the dewatering control module 5, ultra-high pressure dewatering of mineral slurry is achieved. The mineral slurry dewatering mechanism 3 in the integrated dewatering characterization device utilizes the mechanical rigidity of the dual shafts to transmit pressure. Furthermore, the servo driver 33 and servo motor 31 can improve the timely control and high precision requirements of the equipment. The dewatering control module 5 ensures the high precision of the mineral slurry dewatering mechanism 3 and reduces the error of the dewatering characterization data of fine particles. At the same time, the dual-shaft transmission design of the mineral slurry dewatering machine breaks the force limitation of traditional single-shaft pressurization. The two dewatering transmission shafts 35 are symmetrically distributed, which can evenly distribute the pressure to various areas of the pressure dewatering container 6, avoiding the problems of equipment deformation caused by excessive force on one side in traditional equipment.

[0223] It should be noted that, for the sake of simplicity, the foregoing embodiments are all described as a series of actions. However, those skilled in the art should understand that the present invention is not limited to the described order of actions, as some steps may be performed in other orders or simultaneously according to the present invention. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to the present invention.

[0224] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit the scope of protection of the invention. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on these embodiments, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art can still combine, add, delete, or otherwise adjust the features of the various embodiments of the present invention according to the circumstances without conflict or creative effort, thereby obtaining different technical solutions that do not fundamentally depart from the concept of the present invention. These technical solutions also fall within the scope of protection of the present invention.

Claims

1. A high-precision pressure dewatering device for mineral slurry, characterized in that, The device includes: The device body includes a device base and an outer frame, with the outer frame fixedly mounted on the device base. An integrated dewatering and characterization device is installed inside the base of the equipment and is used for pressurized dewatering of slurry; The in-situ drying equipment is installed inside the outer frame of the structure and is used to dry the filter cake after pressure dehydration. The real-time monitoring module is used to collect dehydration-related data in real time during the dehydration and drying process, preprocess the dehydration-related data, and output a real-time status set. The dehydration control module is used to acquire a real-time status set and perform high-precision motion control on the integrated dehydration characterization equipment based on the real-time status set. The integrated dehydration characterization equipment includes: The feeding mechanism is set on the device base and includes a raw material mixing tank, a first gas compression pump and a raw material feeding pipe. The raw material mixing tank is connected to the pressure dehydration container through the raw material feeding pipe, and the pressure dehydration container is also connected to the first gas compression pump. The pressure dewatering operation vessel is set inside the outer frame of the structure and is used to assist in the pressure dewatering treatment of slurry; A slurry dewatering mechanism is installed within the outer frame of the structure and is connected to a pressure dewatering container. The slurry dewatering mechanism is used to pressurize and dewater the slurry within the pressure dewatering container. The slurry dewatering mechanism includes: A servo motor, which is fixedly mounted on the device base; A servo driver is fixedly connected to the output shaft of a servo motor, and the servo driver is electrically connected to the dehydration control module; The worm gear reducer is installed inside the device base and is connected to the servo motor; The dehydration drive shaft is rotatably mounted inside the outer frame of the structure, and is rotatably connected to the worm gear reducer via a transmission belt. A dewatering drive assembly is slidably installed within the outer frame of the structure, and the dewatering drive part is connected to the dewatering drive shaft. The dewatering drive assembly is used to pressurize and dewater the slurry in the pressure dewatering operation container and prevent slag from adsorbing onto the surface; the dewatering drive assembly includes: A pusher sleeve is threaded onto the outer wall of the dehydration drive shaft, and the pusher sleeve is slidably connected to the outer frame of the structure. The dehydration push rod is fixedly connected to the push sleeve; An anti-accumulation pushing part is fixedly connected to the dewatering push rod and is used for pressurizing and dewatering the slurry; The anti-accumulation pushing part includes: A slurry pusher sleeve is detachably installed at the end of the dewatering push rod; A filter cake forming seat is slidably installed in the slurry pushing sleeve. A turbulence guiding groove is provided in the filter cake forming seat. The turbulence guiding groove is used to turbulently blow off the deposits on the surface of the filter cake forming seat to prevent slag from adsorbing on the surface of the filter cake forming seat. An auxiliary turbulence channel is formed on the side wall of the filter cake forming seat and is used to assist the operation of the turbulence guiding channel. A reset buffer spring is fixedly connected to the filter cake forming seat. The reset buffer spring is fixedly embedded in the slurry pushing sleeve, and one end of the reset buffer spring is fixedly connected to the filter cake forming seat. An anti-accumulation protection part is installed inside the slurry pushing sleeve. The anti-accumulation protection part is connected to the filter cake forming seat and is used to assist the filter cake forming seat in resetting and to assist in removing slag from the surface of the filter cake forming seat.

2. The high-precision pressure dewatering device for slurry according to claim 1, characterized in that: The anti-accumulation protection unit includes: The main extrusion seat is slidably disposed within the slurry pushing sleeve, and one end of the main extrusion seat is fixedly connected to the filter cake forming seat; The driven extrusion seat is rotatably installed inside the protective positioning seat, which is fixedly installed in the slurry pushing sleeve. The driven extrusion seat and the main extrusion seat are slidably connected. A torsion spring support is fixedly connected to the driven extrusion seat. The torsion spring support is rotatably connected to the slurry push sleeve. A reset torsion spring is embedded in the torsion spring support. One end of the reset torsion spring is fixedly connected to the protection positioning seat. The spring reset rod is fixedly installed inside the protective positioning seat, and one end of the spring reset rod is fixedly connected to the main compression seat.

3. The high-precision pressure dewatering device for slurry according to claim 2, characterized in that: The in-situ drying equipment includes: A nitrogen cylinder is fixedly installed on the device base, and the outlet of the nitrogen cylinder is connected to the second gas compression pump. A nitrogen heating chamber, wherein the nitrogen heating chamber is used to heat nitrogen, and the nitrogen heating chamber is connected to a second gas compression pump; The nitrogen transmission pipe is used to guide and transmit heated nitrogen gas. One end of the nitrogen transmission pipe is connected to the nitrogen heating box, and the other end is connected to the slurry pushing sleeve. The microwave drying unit is installed inside the pressure dehydration container and is used to perform auxiliary microwave drying on the filter cake after pressure dehydration and shaping.

4. The high-precision pressure dewatering device for slurry according to claim 3, characterized in that: The pressure dehydration operation container includes: A dehydration work stand is connected to a second gas compression pump, and the microwave drying component is disposed inside the dehydration work stand. A push guide seat is fixedly installed on both sides of the dewatering work seat. The push guide seat is slidably connected to the slurry push sleeve and is used to guide and limit the slurry push sleeve. The microwave drying assembly includes: An adjustable motor is fixedly installed inside the dehydration work seat; A first gear is fixedly connected to the output shaft of the regulating motor, and the first gear is rotatably installed inside the dehydration work seat; The second gear, which is sleeved in the dehydration work seat, rotates and meshes with the first gear. A generator mounting seat is fixedly connected to one side of the second gear. A microwave generator, wherein the microwave generator is detachably mounted in the generator mounting base.

5. The high-precision pressure dewatering device for slurry according to claim 4, characterized in that: The real-time monitoring module includes: A multi-dimensional sensor group is used to collect dehydration-related data in real time during the dehydration and drying process. The multi-dimensional sensor group includes a grating displacement sensor, a stress sensor, a vibration monitoring sensor, a laser ranging displacement sensor, a temperature sensor, a liquid level sensor, a weighing sensor, and a capacitive humidity sensor. The preprocessing unit is used to acquire dehydration-related data during the dehydration and drying process, filter and reduce noise on the dehydration-related data during the dehydration and drying process, form a time-series real-time status set, and feed the real-time status set back to the dehydration control module. The dehydration control module includes: The extended state space model module is used to acquire the real-time state set, and construct an extended state space model based on the real-time state set and associated data physical characteristics, time-varying parameters, and external disturbances. The extended state space model is a discrete-time state equation, and outputs the state vector, disturbance estimate, and time-varying parameters based on the extended state space model. The model prediction cross-coupling control optimization module is used to obtain the state vector, disturbance estimate, and time-varying parameters output by the extended state space model. It uses the extended state space model to perform state prediction and error prediction, optimizes the objective function and sets constraints. It uses quadratic programming to transform the objective function into a problem that can be solved using the fast projection gradient method, and obtains the optimized state prediction and error prediction. The neural network compensation module is used to train a neural network compensation model to learn the mapping relationship between state features and error quantities by taking the real-time state features of the system as input and error samples as labels. The neural network compensation model predicts the error quantity based on the current state features, which is then superimposed on the control quantity of the predicted cross-coupling control optimization module to obtain the final control quantity, and the final control quantity is fed back to the servo drive.

6. The high-precision pressure dewatering device for slurry according to claim 5, characterized in that: The neural network compensation model includes a time-series data input layer and a wear level embedding layer set in parallel. The time-series data input layer is used to standardize the preprocessed time-series data. The time-series data input layer is connected to a 1D-CNN local feature extraction layer. The 1D-CNN local feature extraction layer is connected to a Transformer global dependency modeling layer. The Transformer global dependency modeling layer is also connected to an average pooling layer. The wear level embedding layer is connected to a parameter normalization layer. The parameter normalization layer is connected to a fully connected embedding layer. The fully connected embedding layer and the average pooling layer are set in parallel. The fully connected embedding layer and the average pooling layer are respectively connected to a feature fusion layer. A fully connected decision layer is set between the feature fusion layer and the output layer.

7. A high-precision pressure dewatering method for slurry, implemented using the high-precision pressure dewatering device for slurry as described in claim 6, characterized in that: The high-precision pressure dewatering method for the slurry includes: During the feeding stage, the slurry in the raw material mixing tank is stirred, and the slurry is pumped into the dewatering working seat by the first gas compression pump. When the liquid level sensor in the dewatering working seat detects that the slurry liquid level has reached the preset liquid level, the first gas compression pump stops pumping the slurry. During the pressurization and dewatering stage, the dewatering control module sends control parameters to the servo driver. In response to the control parameters, the servo driver starts the servo motor, which drives the dewatering push component to synchronously pressurize the slurry in the dewatering work seat in a dual-axis manner. During pressurization, dewatering-related data is collected in real time by a multi-dimensional sensor group. The dewatering-related data is preprocessed to obtain a real-time state set, which is fed back to the dewatering control module. Based on the real-time state set, a discrete-time extended state space model containing state variables, input variables, and unmodeled dynamic disturbance terms is constructed. The current state vector is obtained in real time, and the system matrix and time-varying parameters are estimated. The extended state space model module generates a control sequence in the future time domain through state prediction and error prediction, and uses a quadratic programming algorithm to continuously optimize and solve for the optimal pressurization pressure and feed flow rate at the current moment to achieve multi-variable coupled control. The error compensation amount between the model prediction value and the actual slurry dewatering effect is calculated in real time through a neural network compensation model, and the compensation amount is fed back to the model prediction cross-coupled control module to correct the current control input, obtain the final control amount, and feed the final control amount back to the servo driver. During the drying stage, after pressurization, the nitrogen cylinder, nitrogen heating box, and microwave drying assembly are started simultaneously. The filter cake is dried by the combined action of hot nitrogen and microwave heating. The temperature inside the dehydration work seat is monitored in real time by a fiber optic temperature sensor, and the dehumidification work seat is dehumidified by a vacuum pump. Drying stops when the humidity sensor in the dehydration work seat detects a humidity of ≤5%.

Images