Horizontal bar machining center oil injection device

Abstract

This invention provides a horizontal strip machining center oil injection device, belonging to the field of machining center equipment technology. It includes a data acquisition and preprocessing module, a collaborative trajectory planning and linkage control module, an intelligent oil injection decision module, a motion and fluid collaborative control module, and a self-optimization and health management module. It constructs a machining state feature vector through multimodal sensor fusion, generates the desired nozzle position using a bidirectional decoupled following strategy, dynamically adjusts the oil injection flow rate, pressure, frequency, and injection angle based on a T-S fuzzy neural network controller, and achieves collaborative control of motion and fluid through a cascaded double closed-loop structure. Simultaneously, it constructs an incremental random forest learning engine for parameter self-optimization and equipment health assessment. This invention achieves precise nozzle position tracking of the tool trajectory, dynamic adaptation of oil injection parameters to machining states, and collaborative operation of oil injection and chip removal, improving cooling and lubrication accuracy and system operational reliability.

Description

Technical Field

[0001] This invention relates to the field of machining center equipment technology, specifically to an oil spraying device for a horizontal strip machining center. Background Technology

[0002] In horizontal bar machining centers, the injection and supply of cutting fluid or lubricating oil is crucial for ensuring machining quality, extending tool life, and timely chip removal during metal cutting. Currently, conventional oil spraying systems typically employ fixed nozzles or single-axis moving nozzles, installed near the machining center spindle or on one side of the worktable. During machining, they spray cooling and lubricating media onto the cutting area at a constant flow rate and pressure. Some systems use timed or quantitative open-loop control, achieving intermittent control of the oil spraying action through the start and stop of solenoid valves or pumps. Chip removal is accomplished by a separate auger or chain-plate chip conveyor, lacking a coordination mechanism with the oil spraying system.

[0003] However, existing oil injection devices have shortcomings in practical applications, such as: the nozzle position is fixed or can only move along a single axis, and cannot be adjusted in real time in three-dimensional space to follow the tool trajectory. When machining complex curved surfaces or performing multi-axis linkage machining, the injection point deviates from the actual cutting area, resulting in poor cooling and lubrication effects and serious oil waste; the oil injection parameters (flow rate, pressure, injection frequency, etc.) are constant and cannot be adaptively adjusted according to dynamic working conditions such as temperature changes, vibration state, and feed rate during the machining process, making it difficult to meet the high flow cooling requirements of roughing and the high precision lubrication requirements of finishing; the oil injection system and the chip removal system operate independently and lack linkage and coordination. During oil injection, the oil may impact the chip removal mechanism and cause splashing, while during chip removal, if the oil volume is too large, it will affect the drying and conveying efficiency of the chips; there is a lack of adaptive and self-diagnostic capabilities for abnormal conditions such as sensor failure and equipment aging, resulting in poor system robustness. After long-term operation, the oil injection accuracy decreases and maintenance relies on manual experience. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides a horizontal strip machining center oil spraying device, which solves the problem.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a horizontal strip machining center spraying device, comprising a moving platform, a control component mounted on the right side of the moving platform, a first servo motor mounted on the outer wall of the front end of the right side of the moving platform, an auger fixedly connected to the output end of the first servo motor, a support frame fixedly connected to the outer wall of the front end of the moving platform, a base fixedly connected to the outer wall of the front end of the moving platform, a machining table mounted on the top of the base, a second servo motor mounted on the top of the rear end of the moving platform, a first linear module fixedly connected to the output end of the second servo motor, first guide seats provided at both ends of the first linear module, and the first guide seats fixedly connected to the moving platform, with a sliding connection on the outer wall of the first guide seats. The sliding block has a sliding platform slidably connected to the top of the first linear module, and the sliding platform is fixedly connected to the sliding block. High-pressure storage tanks are provided at both ends of the sliding platform. A third servo motor is installed inside the sliding platform. The output end of the third servo motor is fixedly connected to the third linear module. A third guide seat is installed inside the sliding platform. An injection mechanism is installed on the top of the third linear module. A second servo motor is fixedly connected to the top of the sliding platform. The output end of the second servo motor is fixedly connected to the second linear module. A second guide seat is provided on the outer wall of both ends of the sliding platform. A moving seat is slidably connected to the outer wall of the second guide seat. A connecting platform is installed on the outer wall of the moving seat. A crankshaft extrusion threaded oil injection pipe is installed on the outer wall of the injection mechanism.

[0006] Preferably, the control unit integrates a data acquisition and preprocessing module, a collaborative trajectory planning and linkage control module, an intelligent fuel injection decision module, a motion and fluid collaborative control module, and a self-optimization and health management module.

[0007] The data acquisition and preprocessing module is used to acquire temperature signals, vibration signals, oil injection flow signals, motor current signals and position feedback signals in the processing area in real time, and to perform filtering, anomaly removal, multimodal fusion and redundancy verification, and output processing status feature vectors.

[0008] The collaborative trajectory planning and linkage control module is used to parse the processing path instructions, calculate the expected position of the fuel injector using a two-way decoupling following strategy, and generate the position instruction sequence of each linear module through S-curve acceleration and deceleration planning. At the same time, it dynamically adjusts the chip removal trigger threshold according to the processing stage and outputs the chip removal trigger signal.

[0009] The intelligent fuel injection decision module is used to dynamically calculate the target fuel injection flow rate, target injection pressure, target injection frequency, and injection angle correction coefficient based on the processing state feature vector, the desired position of the fuel injector, and the processing stage, and corrects the actual fuel injection quantity through the TS fuzzy neural network controller, and outputs the actual fuel injection parameters.

[0010] The motion and fluid coordinated control module is used to receive position command sequences and injection parameter commands. It adopts a cascaded double closed-loop structure to realize the linkage control of the position loop and the flow loop. It also performs coordinated operation of injection and chip removal according to the chip removal trigger signal, outputs servo drive signals and fluid control signals, and feeds back actual operating parameters.

[0011] The self-optimization and health management module receives the processing status feature vector, actual injection parameters, and actual operating parameters. It optimizes the control parameters through an incremental random forest learning engine and sends them back to the intelligent injection decision module. At the same time, it constructs an equipment health assessment model and outputs maintenance prompts and early warning signals.

[0012] This invention provides an oil spraying device for a horizontal strip machining center. It has the following beneficial effects:

[0013] 1. Through the cooperation of the first linear module, the second linear module and the third linear module, the oil injection mechanism can move precisely in the three axes of X, Y and Z (or at least three degrees of freedom) to follow the machining position of the tool or workpiece and adjust the position of the oil nozzle in real time, so as to ensure that the cutting fluid or lubricating oil "precisely hits" the cutting area, avoiding the waste and inadequate cooling of traditional fixed nozzles.

[0014] 2. This invention uses a data acquisition and preprocessing module to perceive multi-dimensional state information such as temperature, vibration, feed speed and motor current in the processing area in real time, constructs a processing state feature vector, and dynamically calculates the target injection flow rate, injection pressure, injection frequency and injection angle based on the TS fuzzy neural network controller, so that the injection parameters can be adaptively adjusted according to the changes in processing conditions, taking into account the high flow cooling requirements in the roughing stage and the high precision lubrication requirements in the finishing stage.

[0015] 3. This invention employs a bidirectional decoupled following strategy through a collaborative trajectory planning and linkage control module, decomposing the desired position of the fuel injector into a following component and a compensation component, thereby achieving precise tracking of the fuel injector position to the tool trajectory. Simultaneously, a time-series collaborative scheduler integrates the fuel injection and chip removal actions into a unified time axis scheduling, dynamically adjusting the chip removal trigger threshold according to the machining stage, and adjusting the fuel injection flow rate and injection angle in linkage during the chip removal process to avoid oil impacting the chip removal mechanism and causing splashing, thus achieving coordinated operation of fuel injection and chip removal.

[0016] 4. This invention employs a cascaded dual-closed-loop structure in the motion and fluid coordinated control module to achieve linkage control between the position loop and the flow loop. In high-speed feed or severe vibration scenarios, a feedforward compensation mechanism is activated to eliminate injection response lag. Simultaneously, an incremental random forest learning engine is constructed through the self-optimization and health management module to continuously optimize the injection control parameters, so that the injection strategy gradually approaches the optimal level with the accumulation of processing, thereby improving the cooling and lubrication accuracy and system reliability under varying operating conditions.

[0017] 5. This invention constructs an equipment health assessment model through a self-optimization and health management module. It comprehensively scores the health status based on multiple dimensions of indicators, such as servo motor current fluctuations, linear module friction changes, sensor drift, and pressure holding capability. It also has abnormal self-healing and graded early warning functions. When a sensor fails, it can automatically switch to indirect estimation mode to maintain basic operation. When the equipment performance deteriorates, it outputs maintenance prompts in a timely manner, effectively reducing the risk of unplanned equipment downtime and extending the service life of the equipment. Attached Figure Description

[0018] Figure 1 This is a schematic diagram of the overall structure of the present invention;

[0019] Figure 2 This is a schematic diagram of the overall rear view of the present invention;

[0020] Figure 3 This is a schematic diagram of the structure of the second guide seat of the present invention;

[0021] Figure 4 This is a schematic diagram of the spray mechanism of the present invention;

[0022] Figure 5 This is a schematic diagram of the control component architecture of the present invention.

[0023] The components include: 1. Mobile platform; 2. Base; 3. Processing table; 4. First servo motor; 41. Screw conveyor; 5. Support frame; 51. Second servo motor; 52. Crankshaft extrusion threaded oil injection pipe; 53. First linear module; 54. First guide seat; 55. Sliding block; 56. Mobile table; 6. Second servo motor; 61. Second linear module; 62. Second guide seat; 63. Mobile seat; 64. Connecting table; 7. High-pressure storage tank; 71. Injection mechanism; 8. Third servo motor; 81. Third linear module; 82. Third guide seat; 89. Control components. Detailed Implementation

[0024] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0025] like Figure 1-4As shown, this embodiment of the invention provides a horizontal strip machining center spraying device, including a mobile platform 1. The device is characterized by: a control component 9 installed on the right side of the mobile platform 1; a first servo motor 4 installed on the outer wall of the front end of the right side of the mobile platform 1; an auger 41 fixedly connected to the output end of the first servo motor 4; a support frame 5 fixedly connected to the outer wall of the front end of the mobile platform 1; a base 2 fixedly connected to the outer wall of the front end of the mobile platform 1; a machining table 3 installed on the top of the base 2; a second servo motor 51 installed on the top of the rear end of the mobile platform 1; a first linear module 53 fixedly connected to the output end of the second servo motor 51; a first guide seat 54 provided at both ends of the first linear module 53; the first guide seat 54 fixedly connected to the mobile platform 1; and a sliding block 55 slidably connected to the outer wall of the first guide seat 54. A movable platform 56 is slidably connected to the top of group 53, and the movable platform 56 is fixedly connected to the sliding block 55. High-pressure storage tanks 7 are provided at both ends of the movable platform 56. A third servo motor 8 is installed inside the movable platform 56. A third linear module 81 is fixedly connected to the output end of the third servo motor 8. A third guide seat 83 is installed inside the movable platform 56. An injection mechanism 71 is installed on the top of the third linear module 81. A second servo motor 6 is fixedly connected to the top of the movable platform 56. A second linear module 61 is fixedly connected to the output end of the second servo motor 6. A second guide seat 62 is provided on the outer wall of both ends of the movable platform 56. A movable seat 63 is slidably connected to the outer wall of the second guide seat 62. A connecting platform 64 is installed on the outer wall of the movable seat 63. A crankshaft extrusion threaded oil injection pipe 52 is installed on the outer wall of the injection mechanism 71.

[0026] The software system of the control component 9 adopts a modular and layered architecture, such as... Figure 5 As shown, the system includes: a data acquisition and preprocessing module, a collaborative trajectory planning and linkage control module, an intelligent fuel injection decision module, a motion and fluid collaborative control module, and a self-optimization and health management module. These modules interact and couple their functions through standardized data interfaces, constructing a complete technical chain of "state perception - trajectory planning - parameter decision - linkage execution - iterative optimization." This enables real-time tracking of the fuel injector position to the tool trajectory, dynamic adaptation of fuel injection parameters to the machining state, and coordinated control of fuel injection actions and chip removal operations, comprehensively improving the cooling and lubrication accuracy and system reliability of the machining center under varying operating conditions.

[0027] The data acquisition and preprocessing module processes data according to an integrated workflow of "acquisition-filtering-fusion-verification", as follows:

[0028] The temperature sensor is connected to the module via an analog signal interface, and reads the temperature signal of area 3 of the processing table in real time at a sampling frequency of 2Hz. Vibration sensor through The interface is connected to collect three axial vibration acceleration signals near the spindle at a sampling frequency of 1Hz. The flow sensor is connected via a pulse signal interface and collects the instantaneous flow signal at the outlet of the injection mechanism 71 at a sampling frequency of 10Hz. Each servo motor driver feeds back its actual position via the CAN bus. ,speed With current The sampling frequency is 100Hz; the grating rulers of each linear module feed back position signals through the SSI interface. The sampling frequency is 100Hz.

[0029] All acquisition channels are equipped with a synchronization timestamp module to ensure time consistency of data from different sensors.

[0030] The signals from each sensor undergo preliminary conversion, transforming voltage / pulse signals into raw physical quantity data: the temperature sensor converts the voltage signal into a temperature value using a built-in calibration coefficient. The vibration sensor converts acceleration signals into vibration amplitude. The flow sensor converts pulse signals into raw flow data by establishing a correlation between the number of pulses and the flow rate. (L / min); The servo motor current is converted to the actual current value through a proportional coefficient. .

[0031] The data acquisition and preprocessing module packages all raw data into a raw data frame (RawData), which includes: temperature data. Vibration data Traffic data Motor current data Location data and .

[0032] The filtering employs an adaptive moving average filtering algorithm to eliminate random noise from the sensor. The filtering formula is as follows:

[0033]

[0034] in The sliding window size is dynamically adjusted based on the signal volatility: for temperature signals... =10, for vibration signals =5, for flow signals =3, for current and position signals =2. Perform filtering operations on each source of raw data to obtain filtered data. .

[0035] The multimodal fusion constructs a processing state feature vector based on the filtered data. The module extracts the spectral features of the vibration signal and calculates the vibration energy spectral density. With main frequency components ; Calculate the rate of temperature change

[0036]

[0037] in =0.5 is the sampling interval. Vibration characteristics, temperature change rate, and feed rate are compared. (Obtained from processing path parsing) and correlated to form a six-dimensional state feature vector.

[0038]

[0039] Used for subsequent decision-making modules.

[0040] Perform signal integrity testing on key sensors. When the flow sensor signal... If three consecutive sampling cycles exceed the normal range or are lost, the module automatically switches to indirect estimation mode, using the motor current. Pressure signal from high-pressure storage tank 7 (Pressure sensor connection) Establish a regression model

[0041]

[0042] in The coefficients were determined through equipment calibration tests, and were taken as follows: =0.12L / (min·A) =0.3L / (min·MPa) =0.08L / min, which enables indirect calculation of fuel injection quantity, ensuring that the system can still maintain basic operation when the sensor fails.

[0043] The data acquisition and preprocessing module will process the feature vector of the processing state. The data is synchronously transmitted to the intelligent fuel injection decision module and the collaborative trajectory planning and linkage control module via a high-speed data bus (CAN / Ethernet), with the data transmission baud rate set to 1Mbps.

[0044] The collaborative trajectory planning and linkage control module receives externally input processing path instructions (G-code format) and state feature vectors sent by the data acquisition module. The process involves injector trajectory planning, tool following control, and chip removal coordination, as detailed below:

[0045] Machining path parsing: The module parses the G-code line by line, extracting the discrete point sequence of the toolpath.

[0046]

[0047] and the corresponding feed rate (mm / s), processing stage indicator (Roughing / Finishing / Idle). A look-ahead algorithm is used during the analysis process to pre-cachise 50 trajectory points for velocity planning and trajectory smoothing.

[0048] Bidirectional decoupling following strategy: The desired position of the fuel injector is decomposed into a following component and a compensation component. Following component

[0049]

[0050] in =(0,50,30)mm is the installation offset vector (x, y, z directions) of the fuel injector relative to the cutting tool, determined by the equipment structural parameters. Compensation component Based on the processing state feature vector Dynamic calculation:

[0051]

[0052] in =0.5 , =0.2 , =0.02s is the compensation coefficient determined through process testing. =50mm / s is the baseline feed rate. Desired position of the fuel injector.

[0053]

[0054] S-curve acceleration / deceleration planning: based on the current position of the fuel injector With desired location The speed of each linear module is generated using the S-curve algorithm. Time curve. Rate of change of acceleration. Maximum acceleration Maximum speed The planning results are output as a sequence of position commands for each axis. .

[0055] Timing Coordination Scheduler: Integrates oil injection and chip removal actions into a unified timeline. The scheduler schedules actions based on the machining stage. Dynamically adjust control strategy:

[0056] Roughing stage: The fuel injector is kept at a large offset from the tool. Each component is increased by 30%, expanding the cooling coverage area; the chip removal trigger threshold is reduced (cumulative fuel injection threshold). =5L, time threshold =15min, to adapt to high cutting volume conditions.

[0057] Finishing stage: Injector offset is reduced to the reference value to improve injection accuracy; chip removal trigger threshold is increased. =15L, =45min), to avoid frequent chip removal interfering with the surface quality of the machined surface.

[0058] Idle phase: The amount of fuel injected is reduced to 20% of the baseline value, and the chip removal action is suspended.

[0059] The chip removal trigger condition is based on a collaborative decision-making process using both the fuel injection accumulation model and the time accumulation model. The module accumulates the total fuel injection volume in real time. With continuous processing time .

[0060] when ≥ or ≥ When the chip removal action is triggered, a chip removal start signal is output to the motion and fluid coordination control module to control the first servo motor 4. Rotation speed drives auger 41 to rotate The oil and chip mixture is pushed to the chip discharge hopper 5 for discharge. After chip discharge is completed, and Reset to zero and restart the accumulation process. If motor current is detected during chip removal... Exceeding the rated value If the value is 150% of 5A (i.e., 7.5A), then chip removal will immediately stop and an alarm signal will be output.

[0061] The collaborative trajectory planning and linkage control module will sequence the position commands for each axis. The chip removal trigger signal is sent to the motion and fluid coordinated control module via a high-speed data bus, and the desired position of the fuel injector is simultaneously determined. Compared with the current processing stage Feedback is sent to the intelligent fuel injection decision module.

[0062] The intelligent fuel injection decision module is based on the processing state feature vector. Fuel injector desired position and processing stage The injection flow rate, pressure, injection frequency, and injection angle are dynamically calculated, as follows:

[0063] Fuzzy Neural Network Controller: Construct a four-input, four-output TS fuzzy neural network model. The input variable is temperature. Vibrational energy Feed rate Depth of cut (Obtained from processing path parsing); the output variable is the target fuel injection flow rate. Target injection pressure Target spray frequency Injection angle correction coefficient (Dimensionless).

[0064] The fuzzy neural network structure comprises 4 input nodes, 12 fuzzy subsets (each input variable is divided into 3 fuzzy sets: low / medium / high or stable / slight / severe), 36 fuzzy rules (initialized through subtractive clustering), and 4 output nodes. The membership function uses a Gaussian function.

[0065]

[0066] in and Here, represents the center and width of the j-th fuzzy set for the i-th input, respectively. The rule consequents use linear functions, and the network parameters are trained using a combination of backpropagation and least squares methods.

[0067] Multi-objective optimization strategy: With optimal processing quality and minimum oil consumption as dual objectives, a weighted evaluation function is constructed.

[0068]

[0069] in =2.0L / min is the maximum fuel injection flow rate. =0.8 Target surface roughness; and These are dynamic weighting coefficients, adjusted according to the processing phase: the roughing stage prioritizes cooling performance. =0.7、 =0.3; the finishing stage focuses on surface quality. =0.3、 =0.7. Predicted surface roughness value. Through the built-in regression model

[0070]

[0071] Calculate the model coefficients. =0.5、 =0.02、 =0.15、 =-0.3、 =-0.1 is determined based on training with historical processing data. The controller satisfies the constraints.

[0072]

[0073]

[0074] Under the premise of J, find the combination of output parameters that minimizes J.

[0075] Adaptive injection pattern adjustment: based on the desired position of the fuel injector Calculate the injection angle correction factor based on the relative geometric relationship with the cutting tool. .

[0076] The module achieves continuous adjustment of the injector orifice diameter by controlling the extrusion stroke of the crankshaft extrusion threaded injection pipe 52. With the angle of injection The relationship is:

[0077]

[0078] in = As the reference angle, = The maximum adjustable range (determined through structural testing) is specified. The extrusion stroke control command is sent to the drive unit of the crankshaft extrusion threaded fuel injection pipe via an analog output interface, enabling real-time adjustment of the injection angle and coverage area.

[0079] Flow closed-loop control: During injection, the flow sensor provides real-time feedback on the actual injection quantity. , and target traffic By comparison, the opening degree of the regulating valve of the high-pressure storage tank (7) or the speed of the pump is corrected using an incremental PID controller. The PID control formula is:

[0080]

[0081] in , (Determined through equipment stability testing).

[0082] The corrected control quantity is sent to the regulating valve drive unit of the high-pressure storage tank 7 through the analog output interface.

[0083] The intelligent fuel injection decision module will determine the target fuel injection flow rate. Target injection pressure Target spray frequency Injection angle control command Data is transmitted to the motion and fluid coordinated control module via a high-speed data bus.

[0084] The motion and fluid coordinated control module receives a sequence of position commands sent by the coordinated trajectory planning module. The system incorporates chip removal trigger signals and injection parameter commands sent by the intelligent injection decision module to achieve unified and coordinated control of servo drive and fluid execution, as detailed below:

[0085] Cascaded double closed-loop structure: the outer loop is the position loop and the inner loop is the flow loop. The two loops are decoupled from each other through a state coupling matrix.

[0086] Position loop control: based on the position commands of each linear module Feedback with grating ruler The deviation is taken as the input, and a feedforward + PID control algorithm is adopted. The feedforward term is the speed planned according to the S-curve. With acceleration Calculate and compensate for the system's inertial delay. The PID control formula is:

[0087]

[0088] in = - (mm), = , = , =50 (Determined through equipment motion stability testing), sampling period =10ms.

[0089] The position loop output is the speed command for each servo motor. The pulse direction is sent to the servo drivers of the first linear module (53), the second linear module (61) and the third linear module (81) through the pulse direction interface.

[0090] Flow loop control: based on the target injection flow rate Feedback from flow sensor The deviation is taken as the input, and an incremental PID controller (with the same parameters as above) is used to calculate the control quantity of the regulating valve opening. .

[0091] Simultaneously, the module receives a feedforward compensation signal, which, in high-speed feed or severe vibration scenarios, adjusts the feed speed accordingly. >300 mm / s or vibration energy > At that time, the feedforward compensation mechanism is automatically activated: the opening of the regulating valve is increased in advance.

[0092]

[0093] in =0.002s / mm =0.1s / (Determined through response characteristic tests) =50mm / s is the base feed rate to eliminate the lag delay in fuel injection response.

[0094] Chip removal linkage control: When a chip removal trigger signal is received, the module executes the following linkage operation:

[0095] Immediately reduce the fuel injection flow rate to the reference value. =20% of 1.0 L / min, that is =0.2L / min, while adjusting the spray angle to the minimum coverage area (extrusion stroke) =0), to prevent oil from directly impacting the screw conveyor 41 and causing splashing;

[0096] Output PWM signal to control the first servo motor 4 Rotation speed drives auger 41 to rotate Real-time monitoring of motor current during chip removal process ,like Then immediately stop and output an alarm;

[0097] After chip removal is completed, the original injection parameters are restored, and a chip removal completion confirmation signal is sent to the collaborative trajectory planning and linkage control module via the high-speed data bus for resetting the cumulative quantity counter.

[0098] The motion and fluid coordinated control module will input the actual operating data (actual speed of each servo motor) Actual location Actual fuel injection quantity Chip removal status The data is synchronously fed back to the self-optimization and health management module via a high-speed data bus.

[0099] The self-optimization and health management module receives operational data from each module, continuously optimizes control parameters, and monitors and predicts the health status of the equipment, as detailed below:

[0100] Incremental Random Forest Learning Engine: Processing State Feature Vectors Actual fuel injection parameters , , and surface quality score of the workpiece after processing (The samples are input by the operator or automatically obtained by the external detection equipment) to build an incremental random forest model.

[0101] The model structure comprises 100 decision trees, each with a maximum depth of 10. The node splitting criterion is minimizing the mean squared error. After a new batch of processing is completed and a surface quality score is obtained, the learning engine performs incremental updates: predicting new samples and calculating the prediction error, selectively updating the decision tree node parameters based on the error magnitude, and simultaneously weighting historical samples to maintain model stability. Every 10 batches of processing, the learning engine evaluates the model performance. If the prediction error continues to decrease, the updated control parameters are transmitted back to the intelligent oil injection decision module via a high-speed data bus, updating the rule weights and membership functions in the fuzzy neural network, allowing the oil injection strategy to gradually approach its optimum with accumulated processing.

[0102] Self-recommended process parameters: When changing workpiece material or cutting tool, the module automatically searches the historical database to find the historical machining record with the highest similarity to the current working condition (material type, cutting tool model, machining requirements), extracts its optimal oil injection parameter set as the initial value, and loads it into the intelligent oil injection decision module. Similarity calculation uses Euclidean distance.

[0103]

[0104] Feature dimensions include material hardness Thermal conductivity Cutting tool diameter Machining accuracy grade G (grades 1–5), depth of cut Weighting coefficient The values ​​are set based on the degree of influence of each feature on the fuel injection parameters. If multiple similar records are found (d < 0.2), the weighted average of the fuel injection parameters is taken as the recommended value, with the weights and similarity scores used in relation to the parameters. Positive correlation.

[0105] Equipment health assessment: The module constructs an equipment health scoring model H (value range 0–100) by integrating the following indicators:

[0106] Servo motor current fluctuation index: ,in =5A, contribution weight ;

[0107] Frictional force variation in linear module: Frictional force is calculated by the deviation between the position loop output command and feedback.

[0108]

[0109] Where m = 20 kg is the mass of the moving part. The contribution weight is calculated using a second-order difference method based on position feedback, taking the moving average value.

[0110] Flow sensor zero drift: Measuring the zero-point offset of the flow sensor under non-processing conditions. Contribution weight ;

[0111] Temperature sensor consistency: Calculating the processing table temperature With ambient temperature Historical trends of differences

[0112]

[0113] Contribution weight ;

[0114] High-pressure storage tank pressure holding capacity: Measure the pressure drop rate within 30 seconds after injection stops. Contribution weight =10 .

[0115] Health score calculation formula:

[0116]

[0117] in This is the current indicator value. This is the baseline value (new equipment status). Allowed range of variation. The baseline values ​​and ranges for each indicator are determined through factory testing. When H is below 60, the module outputs maintenance prompts and displays the specific deviations of each indicator through the human-machine interface.

[0118] Anomaly self-healing mechanism: For non-fatal failures, the module automatically executes compensation strategies.

[0119] Slight flow sensor drift: Online correction of flow measurement values ​​based on historical zero-point drift trends.

[0120]

[0121] in Real-time estimation using the exponentially weighted moving average method;

[0122] Increased noise from the temperature sensor: Automatically increase the filter window size N to 15, while reducing the sampling frequency to 1Hz;

[0123] Increased friction in linear modules: Add an upper limit to the integral term in the position loop PID controller. To prevent integral saturation and improve the feedforward compensation coefficient .

[0124] When the severity of the fault exceeds the self-healing capability, the module outputs a graded warning signal:

[0125] A Level 1 warning (H between 50 and 60) alerts operators to the situation.

[0126] Level 2 warning (H between 40 and 50) limits fuel injection quantity or reduces feed rate;

[0127] A Level 3 fault (H below 40) will immediately shut down the machine and display the fault code.

[0128] The self-optimization and health management module transmits the updated control parameters back to the intelligent fuel injection decision module and the collaborative trajectory planning and linkage control module via a high-speed data bus, and sends the health score and early warning signal to the human-machine interface for operators to monitor in real time.

[0129] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A horizontal strip machining center oil spraying device, comprising a moving platform (1), characterized in that: A control component (9) is installed on the right side of the mobile platform (1). A first servo motor (4) is installed on the outer wall of the front end of the right side of the mobile platform (1). An auger (41) is fixedly connected to the output end of the first servo motor (4). A support frame (5) is fixedly connected to the outer wall of the front end of the mobile platform (1). A base (2) is fixedly connected to the outer wall of the front end of the mobile platform (1). A processing table (3) is installed on the top of the base (2). A second servo motor (51) is installed on the top of the rear end of the mobile platform (1). A first linear module (53) is fixedly connected to the output end of the second servo motor (51). A first guide seat (54) is provided at both the front and rear ends of the first linear module (53). The first guide seat (54) is fixedly connected to the mobile platform (1). A sliding block (55) is slidably connected to the outer wall of the first guide seat (54). A moving table (56) is slidably connected to the top of the first linear module (53). The platform (56) is fixedly connected to the sliding block (55). High-pressure storage tanks (7) are provided at both ends of the platform (56). A third servo motor (8) is installed inside the platform (56). A third linear module (81) is fixedly connected to the output end of the third servo motor (8). A third guide seat (83) is installed inside the platform (56). An injection mechanism (71) is installed on the top of the third linear module (81). A second servo motor (6) is fixedly connected to the top of the platform (56). A second linear module (61) is fixedly connected to the output end of the second servo motor (6). A second guide seat (62) is provided on the outer wall of both ends of the platform (56). A moving seat (63) is slidably connected to the outer wall of the second guide seat (62). A connecting platform (64) is installed on the outer wall of the moving seat (63). A crankshaft extrusion threaded oil injection pipe (52) is installed on the outer wall of the injection mechanism (71).

2. The horizontal strip machining center oil spraying device according to claim 1, characterized in that, The control unit (9) integrates a data acquisition and preprocessing module, a collaborative trajectory planning and linkage control module, an intelligent fuel injection decision module, a motion and fluid collaborative control module, and a self-optimization and health management module. The data acquisition and preprocessing module is used to acquire temperature signals, vibration signals, oil injection flow signals, motor current signals and position feedback signals in the processing area in real time, and to perform filtering, anomaly removal, multimodal fusion and redundancy verification, and output processing status feature vectors. The collaborative trajectory planning and linkage control module is used to parse the processing path instructions, calculate the expected position of the fuel injector using a two-way decoupling following strategy, and generate the position instruction sequence of each linear module through S-curve acceleration and deceleration planning. At the same time, it dynamically adjusts the chip removal trigger threshold according to the processing stage and outputs the chip removal trigger signal. The intelligent fuel injection decision module is used to dynamically calculate the target fuel injection flow rate, target injection pressure, target injection frequency, and injection angle correction coefficient based on the processing state feature vector, the desired position of the fuel injector, and the processing stage, and corrects the actual fuel injection quantity through the TS fuzzy neural network controller, and outputs the actual fuel injection parameters. The motion and fluid coordinated control module is used to receive position command sequences and injection parameter commands. It adopts a cascaded double closed-loop structure to realize the linkage control of the position loop and the flow loop. It also performs coordinated operation of injection and chip removal according to the chip removal trigger signal, outputs servo drive signals and fluid control signals, and feeds back actual operating parameters. The self-optimization and health management module receives the processing status feature vector, actual injection parameters, and actual operating parameters. It optimizes the control parameters through an incremental random forest learning engine and sends them back to the intelligent injection decision module. At the same time, it constructs an equipment health assessment model and outputs maintenance prompts and early warning signals.

3. The horizontal strip machining center oil spraying device according to claim 1, characterized in that, The multimodal fusion and redundancy check include: The temperature signal, vibration signal, oil injection flow signal, servo motor current signal and position feedback signal of each linear module in the processing area are collected in real time synchronously by multiple types of sensors, and a synchronization timestamp is configured for all acquisition channels. After physical quantity conversion of each original signal, an adaptive moving average filtering algorithm is used for filtering, and the size of the filtering window is dynamically adjusted according to the fluctuation characteristics of each signal type. Based on the filtered data, vibration spectrum features and temperature change rate features are extracted, and combined with the feed rate obtained from the processing path analysis, a processing state feature vector representing the current processing state is constructed. The flow sensor performs signal integrity detection. When the flow sensor signal continuously exceeds the normal range or is lost, it automatically switches to indirect estimation mode. A regression model is established by the motor current and the pressure signal of the high-pressure storage tank to achieve indirect estimation of the fuel injection flow. The processing state feature vector is synchronously sent to the intelligent fuel injection decision module and the collaborative trajectory planning and linkage control module via a high-speed data bus.

4. The horizontal strip machining center oil spraying device according to claim 1, characterized in that, The calculation of the desired position of the fuel injector includes: Parse the machining path instructions and extract the discrete point sequence of the tool path and the corresponding feed rate and machining stage identifier; A bidirectional decoupled following strategy is adopted to decompose the desired position of the fuel injector into a following component and a compensation component. The following component is determined based on the installation offset vector of the fuel injector relative to the tool, and the compensation component is dynamically calculated based on the machining state feature vector. The following component and the compensation component are superimposed to generate the desired position of the fuel injector. The installation offset vector is dynamically adjusted according to the processing stage, so that the fuel injector can expand the cooling coverage range in the roughing stage and improve the injection accuracy in the finishing stage.

5. The horizontal strip machining center oil spraying device according to claim 1, characterized in that, The TS fuzzy neural network controller includes: A four-input, four-output TS fuzzy neural network model is constructed. The input variables include temperature, vibration energy, feed rate and depth of cut, and the output variables include target injection flow rate, target injection pressure, target injection frequency and injection angle correction coefficient. A Gaussian function is used as the membership function. Multiple fuzzy sets are divided for each input variable. Fuzzy rules are initialized by subtractive clustering. The rule consequents are linear functions. The network parameters are trained by a combination of backpropagation and least squares. With the dual objectives of optimal processing quality and minimum oil consumption, a weighted evaluation function is constructed, and the weight coefficients are dynamically adjusted according to the processing stage. Under the premise of satisfying the constraints of injection flow rate and injection pressure, the combination of output parameters that minimizes the evaluation function is solved. Based on the relative geometric relationship between the desired position of the fuel injector and the cutting tool, the extrusion stroke of the crankshaft extrusion threaded fuel injector is adjusted by the injection angle correction coefficient to achieve real-time adjustment of the injection angle and coverage range. The actual fuel injection quantity is fed back in real time by a flow sensor. After being compared with the target fuel injection flow, it is corrected in a closed loop by an incremental PID controller. The corrected control quantity is then sent to the regulating valve of the high-pressure storage tank.

6. The horizontal strip machining center oil spraying device according to claim 1, characterized in that, The linkage control includes: It receives the position command sequence and chip removal trigger signal sent by the collaborative trajectory planning module, as well as the injection parameter command sent by the intelligent injection decision module; A cascaded dual closed-loop structure is adopted to realize the linkage control of the position loop and the flow loop. The outer loop is the position loop and the inner loop is the flow loop. The two loops are decoupled from each other through a state coupling matrix. The position loop takes the deviation between the position command of each linear module and the feedback from the grating ruler as input, uses feedforward and PID control algorithms to calculate the speed command of each servo motor, and sends it to the servo driver of each linear module. The flow loop takes the deviation between the target injection flow rate and the feedback from the flow sensor as input and uses an incremental PID controller to calculate the control quantity of the regulating valve opening of the high-pressure storage tank. When the feed rate exceeds the set threshold or the vibration energy exceeds the set threshold, the feedforward compensation mechanism is activated to pre-increase the opening of the regulating valve to eliminate the injection response lag. When a chip removal trigger signal is received, chip removal linkage control is executed: reduce the oil injection flow and adjust the injection angle to the minimum coverage range, output a control signal to drive the chip removal motor to run, monitor the chip removal motor current in real time during the chip removal process, stop chip removal and output an alarm signal when the current exceeds the rated value, and restore the original oil injection parameter settings after the chip removal is completed. The actual operating data is synchronously fed back to the self-optimization and health management module via a high-speed data bus.

7. The horizontal strip machining center oil spraying device according to claim 1, characterized in that, The optimization of the control parameters includes: An incremental random forest learning engine is constructed using the processing state feature vector, actual oil spraying parameters, and workpiece surface quality score after processing as training samples. Once a new batch of samples is processed and a surface quality score is obtained, the incremental random forest learning engine performs incremental updates, predicts new samples and calculates prediction errors, selectively updates decision tree node parameters based on the error magnitude, and performs weighted sampling of historical samples to maintain model stability. After each batch of processing is completed, the model performance is evaluated. If the prediction error continues to decrease, the updated control parameters are sent back to the intelligent fuel injection decision module to update the rule weights and membership functions in the fuzzy neural network. When changing workpiece material or cutting tool, the system automatically searches the historical database to find the historical processing record with the highest similarity to the current working condition, extracts its optimal oil injection parameter set as the initial value and loads it into the intelligent oil injection decision module.

8. The horizontal strip machining center oil spraying device according to claim 1, characterized in that, The construction of the equipment health assessment model includes: Five indicators were considered: servo motor current fluctuation index, linear module friction change, flow sensor zero drift, temperature sensor consistency, and high-pressure storage tank pressure holding capacity. The contribution weight of each indicator was set accordingly. Based on the current value, baseline value, and allowable range of change of each indicator, a weighted calculation is performed to generate an equipment health score; When the health score is lower than the set threshold, maintenance prompts will be output through the human-machine interface and the specific deviation of each indicator will be displayed. For non-fatal failures, an automatic self-healing mechanism is implemented, including online correction of flow measurement values ​​based on zero-point drift trends, automatic increase of filter window size, addition of upper limit limit of integral term in position loop PID controller and increase of feedforward compensation coefficient; When the severity of the fault exceeds the self-healing capability, a graded warning signal is output. Depending on the different intervals of the health score, the following actions are taken: prompting attention, limiting the amount of fuel injection or reducing the feed rate, or immediately stopping the machine and displaying the fault code.

Images