Cutting fluid recycling device for numerical control machine tool

Abstract

The invention relates to the technical field of numerical control machine tool machining, in particular to a cutting fluid recycling device for a numerical control machine tool, which comprises a base, a data acquisition module, a state monitoring module, a separation control module, an adsorption optimization module, a health degree evaluation module and an early warning judgment module, and is characterized in that the data acquisition module is used for acquiring signal parameters; calculating a pollution load index and a stability factor to obtain residual effective time of the cutting fluid; performing centrifugal parameter optimization based on the pollution load index and the residual effective time of the cutting fluid to obtain a separation efficiency coefficient; obtaining an adsorption health index through the separation efficiency coefficient; integrating the parameters to output the health degree of the system; the change rate and the acceleration of the system health degree are analyzed, a comprehensive early warning index is generated, and graded early warning signals are output; multi-dimensional state perception, dynamic parameter optimization and early fault identification are realized, and the operation efficiency, the energy consumption control capability and the predictive maintenance level of the cutting fluid system are improved.

Description

Technical Field

[0001] This invention relates to the field of CNC machine tool processing technology, specifically to a device for recycling cutting fluid for CNC machine tools. Background Technology

[0002] The cutting fluid recycling device for CNC machine tools uses physical, chemical, and biological technologies to recover, purify, and regenerate the used cutting fluid during CNC machine tool processing, and then supplies the cutting fluid that meets the performance standards back to the CNC machine tool, forming a sustainable recycling system.

[0003] In existing technologies, CNC machine tool cutting fluid recycling devices typically rely on single or a few sensor signals, such as only turbidity or flow rate, for condition monitoring and control. This lacks a comprehensive perception of the multi-dimensional state of the cutting fluid system. When faced with complex changes in contaminant composition, fluctuations in colloidal stability, and decay of active ingredients, this isolated monitoring method cannot accurately capture the overall state of the system, leading to inaccurate predictions of the remaining life of the cutting fluid and lag in the adjustment of separation and adsorption parameters. Furthermore, the control strategies of existing CNC machine tool cutting fluid recycling devices are mostly simple feedback control based on fixed thresholds, lacking dynamic modeling and optimization of the coupling relationship between multiple factors such as contaminant load, separation efficiency, and adsorption health. This makes it impossible to achieve adaptive matching between processing parameters and real-time operating conditions, resulting in low system operating efficiency, high energy consumption, and insufficient response to sudden changes in contaminant load.

[0004] In terms of system health management and early warning, existing health assessments are often based on simple linear models or independent indicator judgments, failing to establish nonlinear relationships between key parameters such as pollution load, separation efficiency, adsorption health, and active ingredient concentration, thus failing to provide a systematic health status assessment. Furthermore, existing early warning mechanisms mostly use fixed threshold alarms, failing to comprehensively consider the changing trends and accelerations of health status, and are unable to identify and classify the risks of gradual system degradation or sudden failures in the early stages. This results in early warning information being either too sensitive to produce false alarms or too slow to miss the optimal maintenance opportunity, making it difficult to support the effective implementation of predictive maintenance strategies and increasing the risk of unplanned downtime and maintenance costs. Summary of the Invention

[0005] To address the technical problems raised in the background section, this invention is proposed. An embodiment of this invention provides a cutting fluid recycling device for CNC machine tools.

[0006] The objective of this invention can be achieved through the following technical solutions: A device for recycling cutting fluid for CNC machine tools includes a base, a data acquisition module, a status monitoring module, a separation control module, an adsorption optimization module, a health assessment module, and an early warning judgment module. The device is characterized in that a filtration and purification mechanism is provided at the top of the outer wall of the base, and the data acquisition module is used to collect cutting fluid, particulate matter, and system operating parameters. The condition monitoring module is used to receive cutting fluid, particulate matter and system operating parameters, calculate the pollution load index and stability factor, and obtain the remaining effective time of the cutting fluid; The separation control module optimizes centrifugation parameters and evaluates separation efficiency based on the contamination load index and the remaining effective time of the cutting fluid, and obtains the separation efficiency coefficient. The adsorption optimization module is used to receive the separation efficiency coefficient, optimize the electric field strength, and calculate the adsorption kinetics to obtain the adsorption health index. The health assessment module outputs the system's health status based on the above parameters. The early warning judgment module analyzes the rate of change and acceleration of the system's health status, generates a comprehensive early warning index, and outputs graded early warning signals.

[0007] In a preferred embodiment of the present invention, the filtration and purification mechanism includes a filter box; the bottom of the outer wall of the filter box is fixedly connected to the top of the outer wall of the base; a filter plate is fixedly connected to the inner side wall of the filter box; a motor is fixedly connected to the top of the outer wall of the base via a fixing block; the output end of the motor is provided with a rotating shaft, and one end of the outer wall of the rotating shaft extends into the filter box; a set of cleaning plates is fixedly connected to the outer side wall of the rotating shaft, and the cleaning plates match the filter plates; a set of filter holes are opened at the top of the outer wall of the cleaning plates; an inlet pipe is provided at one end of the outer wall of the filter box; a centrifugal mechanism is provided at one end of the outer wall of the filter box via a connecting pipe; the centrifugal mechanism includes a circular shell; an outlet pipe is provided at one end of the outer wall of the circular shell.

[0008] In a preferred embodiment of the present invention, a circular column is rotatably connected to the top of the outer wall of the base; a centrifuge shell is fixedly connected to the top of the outer wall of the circular column; the centrifuge shell and the filter box are connected by a connecting pipe; the bottom of the outer wall of the circular shell is fixedly connected to the top of the outer wall of the base by an arc plate, and the centrifuge shell and the circular shell are rotatably connected in a sealed manner; a bevel gear one is fixedly connected to the outer wall of the circular column; a rotating rod is rotatably connected to the top of the outer wall of the base by a square block; a bevel gear two is fixedly connected to one end of the outer wall of the rotating rod, and the bevel gear two meshes with the bevel gear one; sprockets are fixedly connected to the outer walls of both the rotating rod and the rotating shaft, and a pair of sprockets are connected by a chain.

[0009] In a preferred embodiment of the present invention, the state monitoring module analysis steps are as follows: Logarithmic normalization is performed on the real-time turbidity value signal of the cutting fluid to compress the high dynamic range; A power-law transform is applied to the submicron particle concentration signal to enhance sensitivity to small particles; The Zeta potential signal on the particle surface is processed by a hyperbolic tangent function to limit the influence of extreme values; An exponential transformation is performed on the real-time dynamic viscosity signal of the cutting fluid to characterize the nonlinear response of the viscosity deviation. The normalized multi-sensor data are synthesized into a pollution load index using an error function probabilistic fusion model. Calculation of Debye length based on ion intensity signal; The colloidal stability factor is calculated by combining the Zeta potential signal and Debye length on the particle surface using an exponential function. The stability factor is corrected using the pollution load index to obtain the corrected stability factor. A performance degradation model is established based on the modified stability factor, and the degradation constant is calculated. By using an exponential decay model and combining it with the running time signal, the remaining effective time of the cutting fluid is predicted.

[0010] In a preferred embodiment of the present invention, the analysis steps for the separation control module are as follows: Based on the pollution load index, a formula for calculating centrifugation time is established to calculate the centrifugation time. The actual separation efficiency is calculated by the concentration difference between the inlet and outlet concentration signals. The theoretical separation efficiency is calculated based on the actual centrifugal force signal, centrifugation time, and real-time dynamic viscosity signal of the cutting fluid using Stokes' sedimentation law. The separation efficiency ratio is calculated by comparing the actual separation efficiency with the theoretical separation efficiency. By considering the overall separation efficiency ratio, processing flow signal, running time, and energy consumption signal, an efficiency coefficient calculation formula is established, and the separation efficiency coefficient is calculated.

[0011] As a preferred embodiment of the present invention, the adsorption optimization module analysis steps are as follows: Based on the separation efficiency coefficient, the residual particle concentration signal, and the real-time dynamic viscosity signal of the cutting fluid, the field intensity modulation factor is calculated through a product model. The field strength is optimized by multiplying the reference field strength by the adjustment factor. The correction factor was calculated using the Henry function. Electrophoretic mobility is calculated using the electrophoretic mobility formula based on the correction factor, the zeta potential signal of the particle surface, the real-time dynamic viscosity signal of the cutting fluid, and the double layer parameters. The macroscopic adsorption rate was calculated by combining electrophoretic mobility, optimized field strength, submicron particle concentration signal and real-time fluid velocity signal; The current adsorption efficiency ratio is calculated by comparing the current adsorption rate with the initial adsorption rate. The flow resistance growth coefficient is calculated by the ratio of the inlet and outlet pressure difference signal to the initial pressure difference. Based on the equipment operating time signal, the time decay factor is calculated using an exponential decay model. The adsorption health index is calculated using a multiplicative model, taking into account the current adsorption efficiency ratio, resistance growth coefficient, and time decay factor.

[0012] In a preferred embodiment of the present invention, the health assessment module analysis steps are as follows: Based on the remaining effective time of the cutting fluid, the weight of the time factor is calculated using an exponentially decaying weight function. Based on the separation efficiency coefficient, the weight of the separation efficiency factor is calculated using the hyperbolic tangent saturation function; Based on the adsorption health index, the weights of adsorption health factors are calculated using an exponential saturation function. Based on the deviation between the real-time active ingredient concentration signal of the cutting fluid and the target concentration, the concentration factor weight is calculated using a Gaussian distribution function. Based on the comparison between the pollution load index and the safety threshold, the weights of the pollution load factors are calculated using a complementary error function; The overall system health score is obtained by aggregating five weighted factors using the geometric mean method.

[0013] In a preferred embodiment of the present invention, the analysis steps of the early warning determination module are as follows: Calculate the first-order difference rate of change of the system health status at consecutive time points; Calculate the second-order difference acceleration of the rate of change in health status; The comprehensive early warning index is obtained by multiplying the exponential amplification term of the degree of health deficiency and the rate of change with the smoothing term of acceleration. Based on the comparison results between the comprehensive early warning index and the dynamic threshold, a graded early warning signal is output, including red alert, yellow alert, blue suggestion and normal status. The dynamic threshold is adjusted periodically over time to adapt to the operational requirements of different time periods.

[0014] In a preferred embodiment of the present invention, the filter box includes a box body and a circular plate; the circular plate and the box body are rotatably connected in a sealed manner; the circular plate and the rotating shaft are fixedly connected; an annular rack is fixedly connected to the outer side wall of the box body; a reciprocating rod is rotatably connected to one end of the inner wall of the cleaning plate, and one end of the outer wall of the reciprocating rod passes through the circular plate; the reciprocating rod and the circular plate are rotatably connected; a reciprocating plate is reciprocally connected to the outer side wall of the reciprocating rod, and the outer side wall of the reciprocating plate is slidably connected to the inner side wall of the cleaning plate; a gear is fixedly connected to one end of the outer wall of the reciprocating rod; the annular rack meshes with a set of gears.

[0015] In a preferred embodiment of the present invention, a square shell is fixedly connected to one side of the outer wall of the filter box; a scraper is fixedly connected to one end of the inner wall of the square shell by a set of springs, and the outer side wall of the scraper is slidably connected to the inner side wall of the square shell; the scraper matches the cleaning plate; baffles are fixedly connected to both sides of the top of the outer wall of the square shell; and a collection box is fixedly connected to one end of the outer wall of the square shell.

[0016] Compared with the prior art, the beneficial effects of the present invention are: 1. The system acquires signals through a data acquisition module; it fuses real-time turbidity, submicron particle concentration, particle surface zeta potential, real-time dynamic viscosity, and ionic strength signals of the cutting fluid to calculate the contamination load index and stability, thus obtaining the remaining effective time of the cutting fluid; it optimizes centrifugation parameters based on the contamination load index and the remaining effective time of the cutting fluid to obtain the separation efficiency coefficient; it receives the separation efficiency coefficient and performs electric field strength optimization and adsorption kinetics calculation to obtain the adsorption health index; it conducts a comprehensive system health assessment based on the remaining effective time of the cutting fluid, the contamination load index, the separation efficiency coefficient, and the adsorption health index to obtain the system health level; based on the system health level, it calculates the first-order differential rate of change and the second-order differential acceleration to obtain the comprehensive early warning index, and performs graded and precise early warning, outputting early warning level signals. This realizes an intelligent management chain from perception, decision-making, execution to assessment, with each module having a single responsibility and clear interfaces, improving system scalability and adaptability under different operating conditions, and enabling multi-dimensional state perception, dynamic parameter optimization, and early fault identification, thereby improving the operating efficiency, energy consumption control capability, and predictive maintenance level of the cutting fluid system.

[0017] 2. By establishing a centrifugation time calculation formula based on the pollution load index, the centrifugation time is output. Since the centrifugation time is adaptively adjusted according to the pollution load index, time and energy are saved when the pollution is light, and the separation effect is guaranteed when the pollution is heavy, thus achieving a balance between efficiency and effectiveness and achieving the goal of energy saving and consumption reduction. Attached Figure Description

[0018] To facilitate understanding by those skilled in the art, the present invention will be further described below with reference to the accompanying drawings.

[0019] Figure 1 This is a structural diagram of the main body of the present invention; Figure 2 This is an exploded structural diagram of the box body and the circular plate of the present invention; Figure 3 This is a structural diagram of the cleaning plate, the ring rack, and the gear of the present invention; Figure 4 This is an exploded view of the reciprocating rod, reciprocating plate, and cleaning plate of the present invention. Figure 5 This is an exploded view of the square shell and scraper of the present invention; Figure 6 This is a structural diagram of the centrifuge mechanism of the present invention; Figure 7 This is an exploded structural diagram of the circular shell and centrifugal shell of the present invention; Figure 8 This is a system block diagram of the present invention; Figure 9 This is a flowchart of the early warning level signal analysis of the present invention; In the diagram: 1. Base; 2. Filtration and purification mechanism; 201. Filter box; 202. Filter plate; 3. Motor; 4. Rotating shaft; 203. Cleaning plate; 204. Filter hole; 5. Liquid inlet pipe; 6. Centrifugation mechanism; 601. Circular shell; 7. Water outlet pipe; 602. Circular column; 603. Centrifuge shell; 8. Connecting pipe; 604. Bevel gear one; 605. Rotating rod; 606. Bevel gear two; 607. Sprocket; 608. Chain; 2011. Box body; 2012. Circular plate; 205. Ring rack three; 206. Reciprocating rod; 207. Reciprocating plate; 208. Gear three; 8. Square shell; 9. Scraper; 10. Baffle; 11. Collection box. Detailed Implementation

[0020] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are also within the scope of protection of the present invention.

[0021] Please see Figures 1-9 As shown, a device for recycling cutting fluid for CNC machine tools includes a base 1, a data acquisition module, a status monitoring module, a separation control module, an adsorption optimization module, a health assessment module, and an early warning judgment module.

[0022] The data acquisition module is used to acquire real-time turbidity value signal, residual particle concentration signal, submicron particle concentration signal, particle surface Zeta potential signal, real-time dynamic viscosity signal, ionic strength signal, processing flow rate signal, real-time active ingredient concentration signal, energy consumption signal, inlet and outlet pressure difference signal, running time signal, actual centrifugal force signal, real-time fluid velocity signal, inlet concentration signal, and outlet concentration signal of the cutting fluid. The status monitoring module is used to receive real-time turbidity value signal, submicron particle concentration signal, particle surface Zeta potential signal, real-time dynamic viscosity signal, ionic strength signal and running time signal of cutting fluid, perform data fusion, calculate the pollution load index and stability of colloidal system, predict the remaining effective life of cutting fluid, and obtain the remaining effective time of cutting fluid. The separation control module optimizes centrifugation parameters and evaluates separation efficiency based on the contamination load index and the remaining effective time of the cutting fluid, and obtains the separation efficiency coefficient. The adsorption optimization module is used to receive the separation efficiency coefficient, optimize the electric field strength, and calculate the adsorption kinetics to obtain the adsorption health index. The health assessment module performs a comprehensive system health assessment based on the remaining effective time of the cutting fluid, the contamination load index, the separation efficiency coefficient, and the adsorption health index to obtain the system health status. The early warning judgment module calculates the first-order differential rate of change and the second-order differential acceleration based on the system's health status. It multiplies the degree of health deficiency, the exponential amplification term of the rate of change, and the smoothing term of the acceleration to obtain a comprehensive early warning index. It then performs graded and precise early warnings and outputs early warning level signals, realizing an intelligent management chain from perception, decision-making, execution to evaluation. Each module has a single responsibility and clear interfaces, which improves the system's maintainability, scalability, and adaptability to different operating conditions.

[0023] Specifically, the analysis steps for the cutting fluid real-time turbidity signal, residual particle concentration signal, submicron particle concentration signal, particle surface Zeta potential signal, cutting fluid real-time dynamic viscosity signal, ionic strength signal, processing flow rate signal, cutting fluid real-time active ingredient concentration signal, energy consumption signal, inlet and outlet pressure difference signal, running time signal, actual centrifugal force signal, fluid real-time flow velocity signal, inlet concentration signal, and outlet concentration signal are as follows: Turbidity data is collected via a turbidity sensor installed in the main circulation pipeline, on a straight section before the filtration system, at a distance of ≥10 times the pipe diameter from the pump outlet and ≥5 times the pipe diameter from the elbow. The measurement range is typically 0-1000 NTU, and the accuracy is typically ±2%FS or ±0.1 NTU. This provides the real-time turbidity signal of the cutting fluid, reflecting the overall concentration of suspended particulate matter. A laser particle size analyzer, installed parallel to the turbidity sensor in the same sampling area, collects data with a concentration range of 0.1-1000 mg / L and a measurement range of 0.1-10 μm, providing the submicron particle concentration signal, reflecting the specific concentration of fine contaminants. A laser particle size analyzer installed in the outlet pipeline of a centrifuge or after a fine filter provides data with an accuracy of ±3% of the reading and a measurement range of 0.1-50 μm, providing the residual particle concentration signal, reflecting the particle concentration after centrifugation or filtration. Finally, a Zeta potential analyzer sensor installed in a dedicated sampling flow path at the branch point of the main circulation pipeline provides data with an accuracy of ±2 mV and a measurement range of -200 to +200 mV. The Zeta potential signal on the particle surface reflects the electrostatic stability of the colloidal system. It is acquired by a viscosity sensor installed in the main circulation pipeline, with an accuracy of ±1% of the reading and a range of 0.1-1000 mPa·s, yielding the real-time dynamic viscosity signal of the cutting fluid, reflecting the internal friction characteristics and compositional changes of the fluid. The TDS sensor, installed in parallel with other sensors in the main circulation pipeline, acquires the signal with a range of 0-200 mS / cm, yielding the ionic strength signal, reflecting the overall electrolyte concentration. An electromagnetic flowmeter installed in the straight pipe section after the main circulation pump acquires the signal with a flow rate range of 0.3-10 m / s and an accuracy of ±0.5% of the reading, yielding the processed flow rate signal, reflecting the system's fluid dynamics. A power transmitter installed in the power control cabinet of the centrifuge mechanism 6 and the supply pump acquires the signal with an accuracy of ±1%FS, yielding the energy consumption signal, reflecting the system's energy consumption and mechanical status. A differential pressure transmitter installed at the high-pressure end (inlet) and the low-pressure end (outlet) acquires the signal with a range of 0-100 kPa and an accuracy of ±0.1%FS, obtaining the inlet and outlet pressure difference signal, reflects the state and contamination load of the cutting fluid. The running time signal, reflecting the cumulative running time of the equipment, is obtained by collecting data through the internal timer of the controller installed in each equipment control unit. The actual centrifugal force signal, reflecting the actual centrifugal force generated by the centrifugal mechanism 6, is obtained by collecting data through a speed sensor and calculation model installed on the circular column 602 of the centrifugal mechanism 6, with an accuracy of ±1% of the reading and a measurement range of 0-10000xg. The turbidity sensor, installed at the inlet pipe 5 and outlet pipe 7, has a pairing accuracy of ±1% relative deviation and a measurement range of 0-1000NTU. The inlet and outlet concentration signals reflect the particle concentrations before and after the treatment equipment. These are collected by electromagnetic flowmeters installed at the inlet of the adsorption unit and the centrifuge mechanism, with an accuracy of ±0.5% of the reading and a measurement range of 0.1-5 m / s, providing real-time fluid velocity signals and reflecting the real-time flow rate within the pipeline. A chemical sensor array installed in the return water pipeline or at monitoring points after the system mixing point measures at a frequency of 1-5 minutes per measurement, obtaining the real-time active ingredient concentration signal of the cutting fluid, reflecting the real-time active ingredient concentration. Multi-sensor fusion provides a comprehensive and three-dimensional data foundation for subsequent analysis, capturing full-dimensional state information from macroscopic turbidity to microscopic potential.

[0024] Specifically, the steps for analyzing the remaining effective time of the cutting fluid are as follows: Based on the real-time turbidity signal of the cutting fluid, the submicron particle concentration signal, the Zeta potential signal of the particle surface, and the real-time dynamic viscosity signal of the cutting fluid, a logarithmic transformation is performed on the real-time turbidity signal of the cutting fluid to handle high dynamic range: ,in This is a turbidity reference value, a constant set based on historical data or system standards, such as the initial turbidity value of new cutting fluid, used for calibration and comparison. The normalized turbidity index is a dimensionless value. A power-law transform is applied to the submicron particle concentration signal to enhance its sensitivity to small particles. ,in This is a reference value for particle concentration, a preset baseline concentration. The normalized particle concentration index is dimensionless. The Zeta potential signal on the particle surface is processed using the hyperbolic tangent function. ,in The normalized Zeta potential exponent is dimensionless. This is the reference value for the Zeta potential. The absolute value of the Zeta potential signal on the particle surface is used to perform an exponential transformation on the real-time dynamic viscosity signal of the cutting fluid: ,in The normalized viscosity deviation index is dimensionless. The reference viscosity is the viscosity value of a new cutting fluid under standard conditions. This is the absolute value of the viscosity deviation, reflecting the viscosity change caused by contamination. The viscosity scaling factor is a constant that determines the rate of change of the exponent. The normalized turbidity index, normalized particle concentration index, normalized Zeta potential index, and normalized viscosity deviation index make the subsequent model more stable and sensitive. A probabilistic fusion model of the error function is established based on the normalized turbidity index, normalized particle concentration index, normalized Zeta potential index, and normalized viscosity deviation index. ,in The error function is a special function in statistics related to the normal distribution. The pollution load index is the final comprehensive index output. To calculate the pollution load index by taking the natural logarithm of the product of the four indices, a stable, dimensionless, and physically meaningful comprehensive index is derived. This index is more accurate than a simple weighted average and can scientifically characterize the overall pollution level of the system. A Debye length formula is established based on the ion intensity signal. ,in For Debye length, ρ is the relative permittivity, describing the solvent's response to an electric field, such as that of water. ≈78, is the vacuum permittivity, with a value of approximately 8.854 × 10⁻⁶. F / m, is the Boltzmann constant, with a value of approximately 1.38 × J / K, The absolute temperature represents the thermodynamic temperature of the system. is Avogadro's constant, with a value of approximately 6.022 × , The elementary charge has a value of approximately 1.602 × A stability index based on energy ratio was established based on the Zeta potential signal and Debye length of the particle surface. ,in Thermal energy is the product of Boltzmann's constant and temperature, representing the energy scale of particle thermal motion. The stability factor represents the stability of a colloidal system. Its value ranges from 0 to 1; the closer the value is to 1, the more stable the system; the closer the value is to 0, the less stable it is. It is based on the DLVO theory. For example, is an exponential function in mathematics, and e is the natural constant, approximately equal to 2.71828. It's another way of writing exponentiation with a base of . A modified stability factor calculation formula is established based on the stability factor and pollution load index. ,in To correct the stability factor, based on Establish a decay model based on instability: ,in The decay constant represents the rate of system performance degradation. It is a dimensionless coefficient used to quantify the speed at which the system's lifetime decreases. A larger value indicates faster system degradation, and a smaller value indicates slower degradation. It is an instability index, representing the degree of instability of the system. The squared instability term amplifies the effect of instability on the decay constant. 0.001 is a scaling factor, an empirical constant, used to adjust the amplitude of the decay constant. An exponential decay model is established based on the decay constant and the running time signal, outputting the remaining effective time of the cutting fluid. The exponential decay model is as follows: ,in This refers to the remaining effective time of the cutting fluid, indicating the amount of time it can still operate normally under the current conditions. The initial life of the cutting fluid is defined by the preset rated total operating time, representing the expected life under ideal conditions, determined based on design specifications, test data, or historical averages. This is an exponential function used to simulate the exponential decay process. Exponential decay is a common lifetime model in engineering, suitable for describing the gradual decline of performance over time. Based on the DLVO theory, the Debye length and stability factor are calculated using the runtime signal, and an exponential decay model incorporating instability is established to make the prediction results more accurate and reliable. The exponential decay model is a proven and effective life prediction method in engineering, which can provide an intuitive remaining effective time and provide a key basis for predictive maintenance, avoiding premature or late replacement of cutting fluid.

[0025] Specifically, the steps for analyzing the separation efficiency coefficient are as follows: A formula for calculating centrifugation time is established based on the pollution load index: ,in The output variable is centrifugation time, which represents the actual time the centrifuge needs to run to achieve the desired separation effect. The baseline centrifugation time is a preset constant representing the basic time required to achieve effective separation under standard operating conditions or light contamination. It is typically determined through extensive experimentation. 500 is a reference baseline value or normalization constant for the contamination load, normalizing the contamination load index. This value is usually derived from the typical contamination range anticipated during design. 0.3 is a time adjustment coefficient or scaling factor, controlling the sensitivity or slope of centrifugation time as the contamination load increases. The calculation represents the percentage increase in time required. The centrifugation time is adaptively adjusted based on the contamination load index, saving time and energy when contamination is light and ensuring separation efficiency when contamination is heavy, thus achieving a balance between efficiency and effectiveness and achieving the goal of energy saving and consumption reduction. A removal rate formula is established based on the inlet and outlet concentration signals. , The inlet concentration is the concentration of contaminants in the cutting fluid before centrifugation. The outlet concentration refers to the concentration of contaminants in the cutting fluid after centrifugation. To determine the actual separation efficiency, an exponential separation efficiency model based on Stokes' sedimentation law is established using actual centrifugal force signals, centrifugation time, and real-time dynamic viscosity signals of the cutting fluid. ,in The theoretical separation efficiency is the optimal separation efficiency that a centrifuge can theoretically achieve under given operating conditions. This is the actual centrifugal force signal. The characteristic diameter of the pollutant particles is known in advance or estimated; 0.0008 is a comprehensive constant, derived and calibrated experimentally or theoretically for a specific system and application. , Establish the efficiency ratio formula: ,in The separation efficiency ratio, also known as the performance ratio or relative efficiency, represents the comparison between the actual performance of a centrifuge and its theoretical best performance under current operating conditions. , The system processes flow and energy consumption signals, establishes an efficiency coefficient calculation formula, and outputs a separate efficiency coefficient. The efficiency coefficient calculation formula is as follows: The comprehensive evaluation indicators take into account effectiveness, efficiency, and cost, comprehensively measure the economics of the separation process, and guide the system towards efficient and low-consumption operation.

[0026] Specifically, the steps for analyzing the adsorption health index are as follows: A multi-factor coupled product model is established based on the separation efficiency coefficient, the residual particle concentration signal, and the Zeta potential signal on the particle surface: ; in The system's preset concentration threshold, used as a warning level, represents a critical safety boundary for water quality. When the water quality approaches or exceeds this value, it means there is a risk of the water quality exceeding the standard. The field-intensity regulating factor is a dimensionless comprehensive adjustment coefficient, with 60 representing the specific field. The empirical normalization constant represents the representative data that can be handled in the system design. The upper limit or turning point, with 1 as the benchmark guarantee value, has a "magnification effect" in its product form. When the efficiency of any upstream link decreases or the current water quality deteriorates, it can significantly increase the electric field strength to compensate, ensuring the stability of the final effluent water quality and making the system highly robust.

[0027] based on Establish a baseline scaling model. ,in By establishing a baseline or setpoint for the system, and through extensive experiments and research, the optimal field strength was identified under standard operating conditions—moderate pollution, moderate Zeta potential, and good upstream operation—to achieve good separation performance with acceptable energy consumption. To optimize the electric field strength, it represents the electric field strength most suitable for the current operating conditions, calculated based on the current system state, such as upstream efficiency, pollution load, and Zeta potential. The formula for electrophoretic mobility is established using the real-time dynamic viscosity signal and correction factor of the cutting fluid. ,in for , This is the real-time dynamic viscosity signal of the cutting fluid. Electrophoretic mobility, representing how fast a particle travels in an electric field, is a key driving parameter for adsorption rate. Let be the relative permittivity of the fluid, a dimensionless physical property parameter determined by the fluid itself, and a known physical constant. ε is the vacuum permittivity, a fundamental physical constant. Here, is the Henry function, and is the correction factor. We can approximate the Henry function using a series of formulas: ,in The dimensionless electric double layer thickness parameter. It is the reciprocal of the Debye length, related to the ionic strength of the solution, and reflects the thickness of the electric double layer. It can be estimated by measuring the conductivity of the solution or preset as an empirical constant. It is the average radius of the pollutant particles, a key design parameter that needs to be obtained through preliminary water quality analysis. It is usually assumed to be a constant, based on... , A macroscopic adsorption flux equation was established using submicron particle concentration signals and real-time fluid velocity signals. ,in The adsorption rate represents the number of pollutants adsorbed onto the electrode per second. This is a submicron particle concentration signal. The effective adsorption area is a system design constant, determined by the geometry, number, and structure of the electrodes. This refers to the real-time fluid velocity signal, i.e., the fluid velocity, based on... Construct the formula for the current adsorption efficiency ratio under the performance retention index: ,in The current adsorption efficiency ratio is a dimensionless ratio used to quantify the degree to which the current actual performance of an adsorption unit is retained relative to its initial novel performance. The initial adsorption rate is a preset constant of the system, representing the adsorption rate that the adsorption unit can achieve under new conditions and standard operating conditions. Based on the inlet and outlet pressure difference signal, the Darcy-Weisbach equation and the formula for the flow resistance growth coefficient under pipeline flow theory are constructed as follows: ,in The relative resistance increase rate, or resistance growth coefficient, is a dimensionless ratio that quantifies the increase in flow resistance of the adsorption unit relative to the initial clean state. This is a signal indicating the pressure difference between import and export points. The initial pressure difference is a preset constant of the system, representing the pressure difference between the inlet and outlet of the adsorption unit when it is in a brand new, thoroughly clean state and operating at rated flow. The time decay factor formula is constructed based on the operating time signal under an exponential time decay model: ,in The time decay factor is a dimensionless coefficient between 0 and 1, representing the degree of performance degradation caused by natural aging of the equipment. The time decay coefficient is an empirical constant that controls the aging rate. It is a natural exponential function with base e, based on the current adsorption efficiency ratio. and Constructing the formula for the adsorption health index under the multifactor multiplicative health index model: ,in The Adsorption Health Index is an indicator used to comprehensively assess the overall health status of the adsorption unit and to quantify the overall health status of the adsorption unit. The weighting coefficient for resistance quantifies the impact of increased flow resistance on the overall health index. This is determined through historical data and experimental calibration. For example, it analyzes historical records to determine what percentage of equipment failures were directly caused by blockages, or simulates different levels of blockage on a test bench to observe their actual impact on the system's final processing performance. This is a benchmark guarantee value, ensuring that the resistance term does not negatively impact the overall health index when no resistance increases. It is introduced to construct the benchmark-penalty model structure and ensure the logical completeness of the formula.

[0028] Specifically, the system health analysis steps are as follows: based on Establish an exponentially decaying time-weighted function model: ,in This is a dimensionless weighting coefficient representing the degree of health degradation caused by system operating time. In the comprehensive health assessment, it reflects the impact of equipment aging on the overall health status. As the benchmark normalization constant, The time decay rate constant is based on Establish a hyperbolic tangent saturated efficiency weight function model: ,in The weighting factor for separation efficiency is a dimensionless weighting coefficient used to quantify the contribution of separation efficiency to the overall health assessment. In the comprehensive health calculation, it represents the health contribution of the centrifugal separation unit's performance. This is a preset constant for the system, representing the expected satisfactory separation performance level that the system can achieve under design or standard operating conditions. As the saturation acceleration coefficient, controlling the saturation rate of the hyperbolic tangent function, an exponential saturation-type health weight function model is established based on the adsorption health index: ,in To assign weights to adsorption health factors, the health index of the adsorption unit is mapped to a dimensionless weighting coefficient in the 0-1 range. In the comprehensive health assessment, this quantifies the contribution of the health status of the adsorption unit to overall health. Using the baseline guaranteed values, we ensure that the lower bound of the weight function's output is 0 and the upper bound approaches 1. We construct a standard 1-exponential decay function to ensure that the function's range conforms to the weight definition. The saturation rate coefficient controls the rate at which the weights increase with the health index. A Gaussian distribution concentration weighting function model is established based on the real-time active ingredient concentration signal of the cutting fluid. ,in The concentration factor weight is a dimensionless weighting coefficient reflecting the degree of matching between the real-time active ingredient concentration signal of the cutting fluid and the target active ingredient concentration of the cutting fluid. In comprehensive health assessment, it quantifies the contribution of concentration control accuracy to overall health. This is the real-time concentration signal of the active ingredients in the cutting fluid. The target cutting fluid active ingredient concentration is a preset optimal concentration value, representing the best operating point in the process design, determined based on product quality standards and process optimization. The constant is the standard form of the Gaussian function, preserving the form of the standard Gaussian probability density function. The relative tolerance coefficient defines the allowable deviation range for concentration control. It is determined based on a comprehensive analysis of the control system's actual ability to maintain concentration stability, process capability, and control accuracy requirements, using historical data as the basis for evaluation. A complementary error function-type pollution weight function model is established based on the pollution load index. ,in This represents the pollution load factor weight, reflecting the risk weight coefficient of the current pollution load relative to the safety threshold. In the comprehensive health assessment, it quantifies the degree of risk impact of the pollution load on overall health. The system presets a safe operating limit as the pollution load threshold, representing the maximum pollution load the system can stably handle. This is a fixed constant set based on the system's processing capacity and safety regulations. For complementary error functions, The warning initiation coefficient defines the warning initiation point for pollution load, which is determined through risk classification analysis and historical fault data. The transition interval coefficient represents the width of the transition interval from safe to dangerous, determined based on the control system's response characteristics and operating time window. , , A geometrically averaged comprehensive health index model is established using concentration factor weights and pollution load factor weights to output the system health score. The geometrically averaged comprehensive health index model includes: This application is in The calculation uses the geometric mean instead of the arithmetic mean because the geometric mean is highly sensitive to the weakest link. Any factor, such as a sharp increase in pollution load or a sudden drop in adsorption health, will lower the overall health score if it is too low, forcing the system to focus on its weakest link. This aligns with the "weakest link" principle in systems engineering, making the calculation of system health more accurate and reliable. Furthermore, this application uses five different functions—exponential, hyperbolic tangent, Gaussian, etc.—to calculate the weights for the five factors. This allows the application to select the most suitable mathematical function for the physical meaning of different factors to capture their dynamic characteristics, ensuring the scientific and rational allocation of weights. For example, the time factor uses an exponential decay function to accurately simulate the natural law of equipment performance decline with increasing operating time; the separation efficiency factor uses... The hyperbolic tangent saturation function effectively reflects the characteristic that the separation efficiency tends to stabilize after reaching a certain level; the adsorption health factor adopts an exponential saturation function, which intuitively shows the process that the contribution of the health status of the adsorption unit to the overall performance gradually saturates as the health index increases; the concentration factor weight is calculated through a Gaussian distribution function to accurately quantify the impact of the matching degree between the active ingredient concentration of the cutting fluid and the target concentration on the system health; the contamination load factor weight uses a complementary error function model to reasonably assess the risk level to the system health after the contamination load exceeds the safety threshold. The diversified function selection strategy significantly improves the accuracy and comprehensiveness of the system health assessment, providing technical support for the stable operation and efficient management of the cutting fluid recycling device for CNC machine tools.

[0029] Specifically, the steps for analyzing the comprehensive early warning index are as follows: Based on system health, a first-order difference rate of change model is established: ,in The health rate reflects the speed at which health changes; a positive value indicates improvement in health, while a negative value indicates deterioration. For the current health status, The health status at the previous moment is used to compare with the current health status and calculate the change. This is a time interval, usually a fixed sampling period, determined according to the needs of the application scenario. For example, if data is sampled once per minute, then... =1 minute, which determines the time unit for calculating the rate of change, based on Establish a second-order difference-based acceleration model: , This represents the acceleration of the rate of change, or the rate of change of the rate of change itself. It reflects whether the rate of change in health status is accelerating or decelerating. A positive value indicates that the rate of change is increasing, and a negative value indicates that the rate of change is decreasing. In time The rate of change, i.e., the rate of change at the current moment. In time The rate of change, i.e., the rate of change at the previous moment, is based on , , Establish a second-order difference-based acceleration model: ,in The comprehensive early warning index is the final quantitative risk indicator used to assess the potential risk level of a system or equipment. The higher the value, the higher the warning level, and the more unstable or dangerous the system is. As for the degree of health deficiency, when When the value approaches 0, this item approaches 1, amplifying the warning index. When the value approaches 1, this item approaches 0, thus lowering the warning index. The exponential function amplifies the effect of the absolute value of the rate of change. The constant 3 is an amplification factor, determined based on experience or experiments, which non-linearly enhances the impact of the rate of change on the early warning index. The larger the rate of change, the faster the exponential term grows. The power function is used to consider the influence of the absolute value of acceleration. The constant 0.5 is a smoothing factor to prevent the warning index from becoming overly sensitive when the acceleration is too large. The greater the acceleration, the greater the contribution of this term, but the growth is relatively slow. This application considers the current state, the rate of change, and the acceleration of change at the same time to realize a three-level warning mechanism. Low health and poor current state, rapid decline in health and large rate of change, and accelerated decline and large acceleration will all lead to a sharp increase in the warning index, effectively distinguishing between chronic deterioration and sudden failure, and realizing more accurate graded warning.

[0030] Specifically, the steps for analyzing early warning level signals are as follows: Based on the comprehensive early warning index The system determines the value of the threshold and outputs signals of different levels. The threshold changes over time, and the system checks the conditions sequentially. Issue a red alert. Output a yellow alert. If the blue suggestions are not met, the normal state is output. Since the warning threshold is not fixed and fluctuates periodically over time, the normal fluctuations caused by periodic changes such as day and night shifts and production rhythm are taken into account to avoid false alarms during fluctuations in normal working conditions, thus improving the reliability of this application.

[0031] This application achieves precise perception, intelligent decision-making, efficient operation, and predictive maintenance of the cutting fluid circulation system through multi-sensor data fusion, a model combining physical mechanisms and data-driven approaches, adaptive parameter optimization, and comprehensive health assessment and dynamic early warning. It has advantages such as extending the life of cutting fluid and equipment, reducing consumable costs and energy consumption, reducing unplanned downtime, and improving product quality and production safety.

[0032] The filtration and purification mechanism 2 includes a filter box 201; the bottom of the outer wall of the filter box 201 is fixed to the top of the outer wall of the base 1; a filter plate 202 is fixed to the inner side wall of the filter box 201; a motor 3 is fixed to the top of the outer wall of the base 1 by a fixing block; the output end of the motor 3 is provided with a rotating shaft 4, and one end of the outer wall of the rotating shaft 4 extends into the filter box 201; a set of cleaning plates 203 is fixed to the outer side wall of the rotating shaft 4, and the cleaning plates 203 match the filter plates 202; a set of filter holes 204 are opened at the top of the outer wall of the cleaning plates 203; an inlet pipe 5 is provided at one end of the outer wall of the filter box 201; a centrifugal mechanism 6 is provided at one end of the outer wall of the filter box 201 through a connecting pipe 8; the centrifugal mechanism 6 includes a circular shell 601; an outlet pipe 7 is provided at one end of the outer wall of the circular shell 601.

[0033] Used cutting fluid enters the filter box 201 through the inlet pipe 5, and is then filtered by the filter plate 202 in the filter box 201. The filtered cutting fluid enters the centrifuge mechanism 6 through the connecting pipe 8 for centrifugal separation and purification. The purified cutting fluid then enters other equipment for further processing or is put into use through the outlet pipe 7. When the cutting fluid is filtered by the filter plate 202, a lot of impurities will adhere to the outer surface of the filter plate 202. At this time, the motor 3 drives the rotating shaft 4 to rotate, and the rotating shaft 4 drives the cleaning plate 203 to rotate, scraping off the impurities on the filter plate 202. The filter holes 204 on the cleaning plate 203 effectively prevent large particles of impurities from entering the subsequent centrifuge mechanism 6, preventing equipment blockage or damage. The motor 3 adopts frequency conversion control technology, which dynamically adjusts the speed according to the accumulation of impurities on the surface of the filter plate 202, ensuring the cleaning effect while reducing energy consumption. The inner wall of the filter box 201 is treated with a nano hydrophobic coating to reduce cutting fluid residue and bacterial growth, and extend the service life of the equipment.

[0034] A circular column 602 is rotatably connected to the top of the outer wall of the base 1; a centrifuge shell 603 is fixedly connected to the top of the outer wall of the circular column 602; the centrifuge shell 603 and the filter box 201 are connected through a connecting pipe 8; the bottom of the outer wall of the circular shell 601 is fixedly connected to the top of the outer wall of the base 1 through an arc plate, and the centrifuge shell 603 is rotatably connected to the circular shell 601 in a sealed manner; a bevel gear 604 is fixedly connected to the outer wall of the circular column 602; a rotating rod 605 is rotatably connected to the top of the outer wall of the base 1 through a square block; a bevel gear 606 is fixedly connected to one end of the outer wall of the rotating rod 605, and the bevel gear 606 meshes with the bevel gear 604; sprockets 607 are fixedly connected to the outer walls of both the rotating rod 605 and the rotating shaft 4, and a pair of sprockets 607 are connected by a chain 608.

[0035] When the rotating shaft 4 rotates, it drives the rotating rod 605 to rotate via the sprocket 607 and chain 608. The rotating rod 605 drives the circular cylinder 602 to rotate via the bevel gear 604 and bevel gear 606. The circular cylinder 602 drives the centrifugal shell 603 to rotate. At this time, the circular shell 601 is in a fixed state. During the high-speed rotation, the centrifugal shell 603 uses centrifugal force to separate the tiny particles and residual impurities in the cutting fluid, resulting in a high purity of the cutting fluid. The sealed rotating connection design between the circular shell 601 and the centrifugal shell 603 effectively prevents cutting fluid leakage and ensures the stability and reliability of the system. The synchronous transmission of power between the rotating rod 605 and the rotating shaft 4 via the sprocket 607 and chain 608 simplifies the system structure and reduces maintenance costs.

[0036] The filter box 201 includes a box body 2011 and a circular plate 2012; the circular plate 2012 and the box body 2011 are sealed and rotatably connected; the circular plate 2012 and the rotating shaft 4 are fixedly connected; an annular rack 205 is fixedly connected to the outer wall of the box body 2011; a reciprocating rod 206 is rotatably connected to one end of the inner wall of the cleaning plate 203, and one end of the outer wall of the reciprocating rod 206 passes through the circular plate 2012; the reciprocating rod 206 and the circular plate 2012 are rotatably connected; a reciprocating plate 207 is reciprocally connected to the outer wall of the reciprocating rod 206, and the outer wall of the reciprocating plate 207 is slidably connected to the inner wall of the cleaning plate 203; a gear 208 is fixedly connected to one end of the outer wall of the reciprocating rod 206; the annular rack 205 meshes with a set of gears 208.

[0037] When the rotating shaft 4 rotates, it drives the circular plate 2012 and a set of reciprocating rods 206 to rotate. The reciprocating rods 206 drive the gear 208 to rotate. Since the gear 208 meshes with the ring rack 205 and the ring rack 205 is fixed, the rotation of the gear 208 causes the ring rack 205 to rotate, thereby driving the reciprocating rod 206 to rotate. The rotation of the reciprocating rod 206 causes the reciprocating plate 207 to reciprocate, sliding back and forth on the inner wall of the cleaning plate 203, thus cleaning the filter holes on the cleaning plate 203. The cleaning process 204 effectively removes fine impurities attached to or accumulated in the cleaning holes, preventing impurities from clogging the filter holes 204 on the cleaning plate 203. This ensures that the filtration and cleaning performance of the filter plate 202 and the cleaning plate 203 is optimal. This application utilizes the meshing transmission principle of gears and ring racks, eliminating the need for an additional power drive device. During the rotation of the cleaning plate 203 driven by the rotating shaft 4, the reciprocating motion of the reciprocating plate 207 is naturally achieved, reducing the energy consumption and manufacturing cost of the equipment and improving the operating efficiency and reliability of the filtration and purification mechanism 2.

[0038] A square shell 8 is fixedly connected to one side of the outer wall of the filter box 201; a scraper 9 is fixedly connected to one end of the inner wall of the square shell 8 by a set of springs, and the outer side wall of the scraper 9 is slidably connected to the inner side wall of the square shell 8; the scraper 9 matches the cleaning plate 203; baffles 10 are fixedly connected to both sides of the top of the outer wall of the square shell 8; a collection box 11 is fixedly connected to one end of the outer wall of the square shell 8.

[0039] When the cleaning plate 203 rotates, when the cleaning plate 203 rotates to the top, the scraper 9 is pressed against one side of the outer wall of the cleaning plate 203 under the action of the spring. As the cleaning plate 203 rotates, it presses the scraper 9, causing the scraper 9 to drive the spring to contract. At this time, the scraper 9 scrapes off the impurities on the cleaning plate 203. The impurities fall into the collection box 11 along with the scraper 9 and the square shell 8 and are collected. The inner wall of the collection box 11 is treated with anti-corrosion to extend its service life and reduce maintenance costs. A damper is provided at the spring. The baffles 10 on both sides of the square shell 8 prevent impurities from splashing out during the scraping process, ensuring a clean and safe working environment.

[0040] In use, this invention filters and purifies the cutting fluid through a filtration and purification mechanism 2 and a centrifugation mechanism 6. However, during purification, some micron-sized particles in the cutting fluid cannot be effectively removed, resulting in a seemingly clean cutting fluid but with diminished processing performance. Furthermore, the cutting fluid itself may become smelly and deteriorate due to the micron-sized particles. Therefore, the data acquisition module, status monitoring module, separation control module, adsorption optimization module, health assessment module, and early warning judgment module of this application detect and analyze the state of the cutting fluid. When the processing performance of the cutting fluid decreases to a certain standard, the early warning judgment module issues an alarm, reminding the operator to replace the cutting fluid promptly. This invention, through the complementary use of these two mechanisms, not only achieves the purification of the cutting fluid... The filtration and purification of macroscopic impurities, combined with an intelligent monitoring and early warning system, captures performance changes in the cutting fluid caused by the accumulation of micron-sized particles. This accurately assesses the current state of the cutting fluid, preventing potential risks such as unstable machining quality and accelerated equipment wear caused by deteriorating cutting fluid performance. Furthermore, data analysis and feedback provide a scientific basis for the optimized use and management of the cutting fluid, forming a closed-loop cutting fluid recycling solution. This not only improves the cleanliness and machining performance of the cutting fluid and extends the service life of both the cutting fluid and the equipment, but also reduces unplanned downtime through precise early warning and timely maintenance, thereby improving production efficiency and product quality. This provides a strong guarantee for the efficient and stable operation of CNC machine tools.

[0041] It should be understood that although the steps in the flowcharts of the various embodiments of the present invention are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the various embodiments may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least a portion of the sub-steps or stages of other steps.

[0042] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0043] The foregoing description is illustrative of the invention and should not be construed as limiting it. Although several exemplary embodiments of the invention have been described, those skilled in the art will readily understand that many modifications can be made to the exemplary embodiments without departing from the novel teachings and advantages of the invention. Therefore, all such modifications are intended to be included within the scope of the invention as defined in the claims. It should be understood that the foregoing description is illustrative of the invention and should not be construed as limiting it to the specific embodiments disclosed, and modifications to the disclosed embodiments and other embodiments are intended to be included within the scope of the appended claims. The invention is defined by the claims and their equivalents.

Claims

1. A device for recycling cutting fluid for CNC machine tools, comprising a base (1), a data acquisition module, a status monitoring module, a separation control module, an adsorption optimization module, a health assessment module, and an early warning judgment module, characterized in that, The outer wall of the base (1) is provided with a filtration and purification mechanism (2), and the data acquisition module is used to collect cutting fluid, particulate matter and system operating parameters; The condition monitoring module is used to receive cutting fluid, particulate matter and system operating parameters, calculate the pollution load index and stability factor, and obtain the remaining effective time of the cutting fluid; The separation control module optimizes centrifugation parameters and evaluates separation efficiency based on the contamination load index and the remaining effective time of the cutting fluid, and obtains the separation efficiency coefficient. The adsorption optimization module is used to receive the separation efficiency coefficient, optimize the electric field strength, and calculate the adsorption kinetics to obtain the adsorption health index. The health assessment module outputs the system's health status based on the above parameters. The early warning judgment module analyzes the rate of change and acceleration of the system's health status, generates a comprehensive early warning index, and outputs graded early warning signals.

2. The cutting fluid recycling device for CNC machine tools according to claim 1, characterized in that, The filtration and purification mechanism (2) includes a filter box (201); the bottom of the outer wall of the filter box (201) is fixedly connected to the top of the outer wall of the base (1); a filter plate (202) is fixedly connected to the inner side wall of the filter box (201); a motor (3) is fixedly connected to the top of the outer wall of the base (1) by a fixing block; the output end of the motor (3) is provided with a rotating shaft (4), and one end of the outer wall of the rotating shaft (4) extends into the filter box (201); the outer side wall of the rotating shaft (4) is fixedly connected to... There is a set of cleaning plates (203), and the cleaning plates (203) are matched with the filter plates (202); a set of filter holes (204) are opened at the top of the outer wall of the cleaning plate (203); an inlet pipe (5) is provided at one end of the outer wall of the filter box (201); a centrifugal mechanism (6) is provided at one end of the outer wall of the filter box (201) through a connecting pipe (8); the centrifugal mechanism (6) includes a round shell (601); an outlet pipe (7) is provided at one end of the outer wall of the round shell (601).

3. A cutting fluid recycling device for CNC machine tools according to claim 2, characterized in that, A circular column (602) is rotatably connected to the top of the outer wall of the base (1); a centrifuge shell (603) is fixedly connected to the top of the outer wall of the circular column (602); the centrifuge shell (603) and the filter box (201) are connected through a connecting pipe (8); the bottom of the outer wall of the circular shell (601) is fixedly connected to the top of the outer wall of the base (1) through an arc plate, and the centrifuge shell (603) and the circular shell (601) are rotatably connected in a sealed manner; the outer wall of the circular column (602 ...). A bevel gear 1 (604) is fixedly connected to the side wall; a rotating rod (605) is rotatably connected to the top of the outer wall of the base (1) through a square block; a bevel gear 2 (606) is fixedly connected to one end of the outer wall of the rotating rod (605), and the bevel gear 2 (606) meshes with the bevel gear 1 (604); a sprocket (607) is fixedly connected to the outer wall of both the rotating rod (605) and the rotating shaft (4), and a pair of sprockets (607) are connected by a chain (608).

4. A cutting fluid recycling device for CNC machine tools according to claim 1, characterized in that, The status monitoring module analysis steps are as follows: Logarithmic normalization is performed on the real-time turbidity value signal of the cutting fluid to compress the high dynamic range; A power-law transform is applied to the submicron particle concentration signal to enhance sensitivity to small particles; The Zeta potential signal on the particle surface is processed by a hyperbolic tangent function to limit the influence of extreme values; An exponential transformation is performed on the real-time dynamic viscosity signal of the cutting fluid to characterize the nonlinear response of the viscosity deviation. The normalized multi-sensor data are synthesized into a pollution load index using an error function probabilistic fusion model. Calculation of Debye length based on ion intensity signal; The colloidal stability factor is calculated by combining the Zeta potential signal and Debye length on the particle surface using an exponential function. The stability factor is corrected using the pollution load index to obtain the corrected stability factor. A performance degradation model is established based on the modified stability factor, and the degradation constant is calculated. By using an exponential decay model and combining it with the running time signal, the remaining effective time of the cutting fluid is predicted.

5. A cutting fluid recycling device for CNC machine tools according to claim 4, characterized in that, The analysis steps for the separation control module are as follows: Based on the pollution load index, a formula for calculating centrifugation time is established to calculate the centrifugation time. The actual separation efficiency is calculated by the concentration difference between the inlet and outlet concentration signals. The theoretical separation efficiency is calculated based on the actual centrifugal force signal, centrifugation time, and real-time dynamic viscosity signal of the cutting fluid using Stokes' sedimentation law. The separation efficiency ratio is calculated by comparing the actual separation efficiency with the theoretical separation efficiency. By considering the overall separation efficiency ratio, processing flow signal, running time, and energy consumption signal, an efficiency coefficient calculation formula is established, and the separation efficiency coefficient is calculated.

6. A cutting fluid recycling device for CNC machine tools according to claim 5, characterized in that, The adsorption optimization module analysis steps are as follows: Based on the separation efficiency coefficient, the residual particle concentration signal, and the real-time dynamic viscosity signal of the cutting fluid, the field intensity modulation factor is calculated through a product model. The field strength is optimized by multiplying the reference field strength by the adjustment factor. The correction factor was calculated using the Henry function. Electrophoretic mobility is calculated using the electrophoretic mobility formula based on the correction factor, the zeta potential signal of the particle surface, the real-time dynamic viscosity signal of the cutting fluid, and the double layer parameters. The macroscopic adsorption rate was calculated by combining electrophoretic mobility, optimized field strength, submicron particle concentration signal and real-time fluid velocity signal; The current adsorption efficiency ratio is calculated by comparing the current adsorption rate with the initial adsorption rate. The flow resistance growth coefficient is calculated by the ratio of the inlet and outlet pressure difference signal to the initial pressure difference. Based on the equipment operating time signal, the time decay factor is calculated using an exponential decay model. The adsorption health index is calculated using a multiplicative model, taking into account the current adsorption efficiency ratio, resistance growth coefficient, and time decay factor.

7. A cutting fluid recycling device for CNC machine tools according to claim 6, characterized in that, The health assessment module analysis steps are as follows: Based on the remaining effective time of the cutting fluid, the weight of the time factor is calculated using an exponentially decaying weight function. Based on the separation efficiency coefficient, the weight of the separation efficiency factor is calculated using the hyperbolic tangent saturation function; Based on the adsorption health index, the weights of adsorption health factors are calculated using an exponential saturation function. Based on the deviation between the real-time active ingredient concentration signal of the cutting fluid and the target concentration, the concentration factor weight is calculated using a Gaussian distribution function. Based on the comparison between the pollution load index and the safety threshold, the weights of the pollution load factors are calculated using a complementary error function; The overall system health score is obtained by aggregating five weighted factors using the geometric mean method.

8. A cutting fluid recycling device for CNC machine tools according to claim 7, characterized in that, The analysis steps of the early warning determination module are as follows: Calculate the first-order difference rate of change of the system health status at consecutive time points; Calculate the second-order difference acceleration of the rate of change in health status; The comprehensive early warning index is obtained by multiplying the exponential amplification term of the degree of health deficiency and the rate of change with the smoothing term of acceleration. Based on the comparison results between the comprehensive early warning index and the dynamic threshold, a graded early warning signal is output, including red alert, yellow alert, blue suggestion and normal status. The dynamic threshold is adjusted periodically over time to adapt to the operational requirements of different time periods.

9. A cutting fluid recycling device for CNC machine tools according to claim 2, characterized in that, The filter box (201) includes a box body (2011) and a circular plate (2012); the circular plate (2012) and the box body (2011) are sealed and rotatably connected; the circular plate (2012) and the rotating shaft (4) are fixedly connected; an annular rack (205) is fixedly connected to the outer wall of the box body (2011); a reciprocating rod (206) is rotatably connected to one end of the inner wall of the cleaning plate (203), and one end of the outer wall of the reciprocating rod (206) is... A circular plate (2012) is passed through; the reciprocating rod (206) and the circular plate (2012) are rotatably connected; a reciprocating plate (207) is reciprocally connected to the outer wall of the reciprocating rod (206), and the outer wall of the reciprocating plate (207) is slidably connected to the inner wall of the cleaning plate (203); a gear three (208) is fixedly connected to one end of the outer wall of the reciprocating rod (206); the annular rack three (205) meshes with a set of gear three (208).

10. A cutting fluid recycling device for CNC machine tools according to claim 9, characterized in that, A square shell (8) is fixed to one side of the outer wall of the filter box (201); a scraper (9) is fixed to one end of the inner wall of the square shell (8) by a set of springs, and the outer side wall of the scraper (9) is slidably connected to the inner side wall of the square shell (8); the scraper (9) matches the cleaning plate (203); baffles (10) are fixed to both sides of the top of the outer wall of the square shell (8); a collection box (11) is fixed to one end of the outer wall of the square shell (8).

Images