High-precision cutting process for woodworking machines based on laser positioning

Abstract

The present application relates to the technical field of wood processing, and discloses a high-precision cutting process for woodworking machinery based on laser positioning, comprising the following steps: S1, scanning the surface of wood by a laser sensor to obtain wood data, S2, combining visible light laser and infrared laser, S3, analyzing the characteristics of wood by using an AI algorithm, S4, dynamic compensation of laser error, S5, multi-process collaborative cutting, S6, cooling and heat prevention technology, S7, intelligent environment regulation: real-time monitoring of factory environment parameters to automatically adjust the working state of the laser cutting system, S8, adaptive laser focal length adjustment, S9, adopting an automatic feedback mechanism to record errors in each cutting process in real time. By adopting dual-spectrum laser fusion ranging technology, combining visible light and infrared laser, precise ranging of different wood materials and surface conditions is realized, and the ranging effect is more adaptive and less error, thereby improving the cutting positioning precision.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of wood processing, specifically to a high-precision cutting process for woodworking machinery based on laser positioning. BACKGROUND

[0002] In the modern wood processing industry, high-precision cutting technology is the key to improving processing efficiency and product quality. Among them, laser ranging as a key positioning method before cutting directly affects the accuracy of the cutting path. However, traditional laser ranging technology mostly relies on single-wavelength laser for measurement, which has poor adaptability to different types of wood and is difficult to meet the high-precision requirements in complex wood environments. For example, factors such as wood material, color, texture, and surface roughness can affect the reflection characteristics of the laser, resulting in ranging errors.

[0003] Part of the wood may affect the absorption and scattering of light due to different surface moisture contents, making it difficult for the ranging system to work stably. Especially in high-precision cutting application scenarios, the accumulation of ranging errors will directly affect the cutting quality, and even cause material waste and production efficiency decline. How to improve the adaptability of the laser ranging system to different types of wood and ensure accurate ranging under complex surface conditions has become an important challenge for the development of intelligent wood processing technology. SUMMARY

[0004] In view of the shortcomings of the prior art, the present application provides a high-precision cutting process for woodworking machinery based on laser positioning, which solves the problem of poor adaptability of single-wavelength laser ranging to different materials and surface characteristics in the prior art, which is prone to ranging errors due to changes in reflectivity or surface texture, and improves cutting precision.

[0005] To achieve the above purpose, the present application realizes the following technical scheme: a high-precision cutting process for woodworking machinery based on laser positioning, comprising the following steps:

[0006] S1, laser positioning technology application: scan the wood surface through a laser sensor to obtain wood data, and then use laser measurement to obtain preliminary positioning information;

[0007] S2, dual-spectrum laser fusion ranging: adopt dual-spectrum laser fusion technology, combine visible light laser and infrared laser, and adjust the wavelength and power of the laser in real time to maintain ranging accuracy;

[0008] S3, AI intelligent path optimization: use AI algorithms to analyze the characteristics of the wood, dynamically optimize the cutting path according to the wood structure characteristics, and adjust the cutting sequence and process according to the real-time scanning data of the wood before each cutting;

[0009] S4, laser error dynamic compensation: during the cutting process, the laser system is used to monitor the error caused by vibration, temperature and humidity change or equipment offset in real time, and IMU inertial sensor and deep learning algorithm are used to dynamically compensate the positioning error;

[0010] S5, multi-process collaborative cutting: by adopting double-process cutting method, pre-scribing is carried out by low-power laser scribing during processing, which is used to define the cutting path, and then CNC mechanical cutting equipment is used for accurate cutting;

[0011] S6, cooling and heat prevention technology: in the process of laser cutting, high-pressure airflow or water mist spraying device is equipped to reduce the temperature of the cutting area and avoid the adverse effects of heat accumulation on the wood surface;

[0012] S7, intelligent environment adjustment: real-time monitoring of factory environment parameters automatically adjusts the working state of the laser cutting system, and automatically adjusts the dust suction intensity according to the dust concentration through the negative pressure dust collection system to prevent dust from interfering with the laser positioning system;

[0013] S8, adaptive laser focal length adjustment: adaptive laser focal length adjustment mechanism is adopted, the thickness change of wood is detected in real time, the laser focal length is automatically adjusted, the laser cutting always maintains the best focal point, and the cutting precision is improved;

[0014] S9, automatic feedback and optimization learning: automatic feedback mechanism is adopted, the error in each cutting process is recorded in real time, cutting parameters are continuously optimized through reinforcement learning algorithm, the best cutting path and parameter adjustment strategy are self-learned, and cutting precision and efficiency are improved with the increase of use time.

[0015] Preferably, in S1, the wood data is used to obtain three-dimensional surface data of the wood based on laser reflection principle to create a digital model of the wood, the laser measurement frequency is set to 2000 times / s, the positioning error is controlled to be ±0.08mm, the laser sensor adopts line laser sensor and plane laser sensor, the line laser sensor scanning frequency is 1000-1200Hz, the measurement range is 0-80m, the accuracy is ±1mm, which is used to obtain the linear profile data of the wood surface, and the plane laser sensor scanning rate is 45000-50000 points / s, the measurement accuracy is ±2mm, which is used to obtain the two-dimensional plane data of the wood surface.

[0016] Preferably, in S2, the visible light laser uses a semiconductor laser source with a wavelength of 400-450 nm and a power range of 5-10 mW, which is used to clearly capture the detailed information of the wood surface, including texture and color detail information, and the infrared laser uses an infrared laser source with a wavelength of 800-940 nm and a power range of 10-20 mW, which is used to deal with the case where the wood surface color is deep or the reflectivity is low, and the real-time adjustment of the wavelength and power of the laser is completed within 45-50 ms by the built-in photoelectric adjustment module according to the feedback information of the wood surface, including reflected light intensity and color.

[0017] Preferably, in S3, the features include texture, knots, cracks, and uneven density, and the optimized cutting path is used to avoid the defect area to avoid wasting wood, and the AI algorithm uses a U-Net neural network algorithm based on deep learning, which has a neural network structure including 5 convolutional layers with convolution kernel sizes of 3x3, 5x5, and 7x7, a step size of 1, and 3 fully connected layers with neuron counts of 128, 64, and 32, respectively, for analyzing the features of wood texture, knots, cracks, and uneven density.

[0018] Preferably, in S4, the IMU inertial sensor is used to monitor the acceleration and angular velocity changes of the device in X, Y, and Z directions in real time, with an acceleration measurement range of ±16g and an accuracy of ±0.01g, and an angular velocity measurement range of ±2000dps and an accuracy of ±0.1dps, and the deep learning algorithm uses a long short-term memory network LSTM algorithm to analyze the data collected by the IMU sensor and the positioning data fed back by the laser positioning system, and to build an error prediction model, which dynamically compensates for the positioning error within 140-150 ms, improving the cutting path accuracy to within ±0.08mm.

[0019] Preferably, in S5, the double-process cutting includes laser pre-scribing and mechanical cutting, the low-power laser pre-scribing has a laser power of 3-5 mW and a laser beam diameter of 0.1-0.2 mm, and a pre-scribing speed of 10-15 mm / s, and the CNC mechanical cutting equipment has a cutting speed of 50-150 mm / s and a cutting accuracy of ±0.03 mm.

[0020] Preferably, in S6, the high-pressure gas jet device has a gas pressure range of 0.5-0.8 MPa and a gas flow rate of 2-3 m 3The water mist spraying device is provided with a nozzle of SV-10 type, the diameter of the nozzle is 2-3 mm, the water mist particle diameter of the water mist spraying device is 10-20 microns, the water spraying pressure is 0.3-0.5 MPa, the water spraying flow is 1-2 L / min, and the fine water mist is formed by the nozzle of WS-20 type, and the number of the nozzle holes is 10-15.

[0021] Preferably, the environmental parameters in S7 include temperature, humidity and dust concentration, the accuracy of the temperature monitoring is ±0.2 DEG C, the accuracy of the humidity monitoring is ±2% RH, and the accuracy of the dust concentration monitoring is ±1 mg / m 3 When the dust concentration is lower than 50 mg / m 3 , the dust suction intensity is adjusted to 200-300 Pa, and when the dust concentration is higher than 80 mg / m 3 , the dust suction intensity is adjusted to 400-500 Pa.

[0022] Preferably, in S8, the wood thickness change is detected in real time, a thickness sensor is used, the detection accuracy is ±0.05 mm, and the detection frequency is 450-500 Hz, and the automatic adjustment of the laser focal length adopts an electric zoom lens.

[0023] Preferably, in S9, the automatic feedback mechanism is used to record error data in the cutting process in real time, the error data includes positioning error, cutting path deviation and cutting depth error, the recording frequency is 10 times per cutting process, the reinforcement learning algorithm adopts a deep Q network (DQN) algorithm, after 20 times of cutting are completed, the cutting parameters are optimized within 300-400 ms according to the recorded error data, the cutting error gradually decreases at a speed of 0.02 mm per optimization, and the cutting efficiency gradually increases at a speed of 2% per optimization.

[0024] The application provides a high-precision cutting process for woodworking machinery based on laser positioning.

[0025] 1. The application adopts dual-spectrum laser fusion ranging technology, combines visible light and infrared laser to achieve accurate ranging of different wood materials and surface conditions, and has stronger adaptability and smaller error, thereby improving cutting positioning accuracy.

[0026] 2. The application combines AI intelligent path optimization, accurately identifies wood texture, knots, scars and cracks based on a U-Net deep learning algorithm, realizes intelligent planning of the cutting path, avoids the flaw area, and improves wood utilization and reduces cutting loss.

[0027] 3、The present application realizes the effects of improving cutting surface quality and more stable precision by using high-pressure airflow and water mist jet cooling system to dynamically control temperature during cutting process, reducing heat-affected zone, and reducing wood burning and deformation.

[0028] 4、The present application realizes the effects of gradually improving precision and stably optimizing cutting quality by using automatic feedback and reinforcement learning optimization mechanism to record error data in real time and optimize cutting parameters based on DQN algorithm to continuously reduce error and improve processing efficiency. BRIEF DESCRIPTION OF DRAWINGS

[0029] Figure 1 The process flow chart of the high-precision cutting process of the woodworking machinery based on laser positioning of the present application. DETAILED DESCRIPTION

[0030] The technical solutions of the present application will be described in detail below with reference to the drawings of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0031] Please refer to the drawings of the present application Figure 1 The embodiment of the present application provides a high-precision cutting process of woodworking machinery based on laser positioning, which comprises the following steps:

[0032] S1, application of laser positioning technology: scanning the surface of wood by laser sensor to obtain wood data, and then using laser measurement to obtain preliminary positioning information;

[0033] S2, dual-spectrum laser fusion ranging: using dual-spectrum laser fusion technology, combining visible light laser and infrared laser, and adjusting the wavelength and power of laser in real time to maintain ranging accuracy;

[0034] S3, AI intelligent path optimization: using AI algorithm to analyze the characteristics of wood, dynamically optimizing the cutting path according to the structure characteristics of wood, and adjusting the cutting order and process according to the real-time scanning data of wood before each cutting;

[0035] S4, laser error dynamic compensation: using laser system to monitor errors caused by vibration, temperature and humidity changes or equipment deviation in real time during cutting process, and using IMU inertial sensor and deep learning algorithm to dynamically compensate positioning errors;

[0036] S5, multi-process collaborative cutting: using double-process cutting method, pre-marking by low-power laser marking during processing, and then using CNC mechanical cutting equipment for accurate cutting;

[0037] S6, Cooling and heat protection technology: During laser cutting, high-pressure airflow or water mist spraying devices are equipped to reduce the temperature of the cutting area and prevent heat accumulation from affecting the wood surface;

[0038] S7, Intelligent environmental adjustment: Real-time monitoring of factory environmental parameters automatically adjusts the working state of the laser cutting system. The dust suction force is automatically adjusted according to the dust concentration through the negative pressure dust suction system to prevent dust from interfering with the laser positioning system;

[0039] S8, Adaptive laser focal length adjustment: Adopting an adaptive laser focal length adjustment mechanism, the laser focal length is automatically adjusted by real-time detection of wood thickness changes, so that the laser cutting always maintains the best focal point, improving cutting precision;

[0040] S9, Automatic feedback and optimization learning: Adopting an automatic feedback mechanism, real-time recording of errors during each cutting process, continuously optimizing cutting parameters through reinforcement learning algorithm, self-learning optimal cutting path and parameter adjustment strategy, improving cutting precision and efficiency with increasing use time.

[0041] In S1, wood data is used to obtain three-dimensional surface data of wood based on laser reflection principle to create a digital model of wood. The laser measurement frequency is set to 2000 times / second, the positioning error is controlled to ±0.08mm, the laser sensor adopts line laser sensor and plane laser sensor, the line laser sensor scanning frequency is 1000-1200Hz, the measurement range is 0-80m, the accuracy is ±1mm, which is used to obtain the linear profile data of the wood surface, the plane laser sensor scanning rate is 45000-50000 points / second, the measurement accuracy is ±2mm, which is used to obtain the two-dimensional plane data of the wood surface.

[0042] Specifically, combined with multi-sensor three-dimensional positioning, dual-spectrum laser ranging, AI intelligent optimization and error dynamic compensation, high-precision wood cutting is realized. A three-dimensional model of wood is constructed by line laser and plane laser sensors, with a measurement frequency of 2000 times / second and a positioning error controlled within ±0.08mm. Visible light and infrared light fusion ranging is adopted, and the laser wavelength and power are dynamically adjusted within 50ms to ensure ranging accuracy. AI intelligent optimization path, based on U-Net algorithm to identify wood texture, knots, and cracks, improves utilization and reduces waste. IMU sensor combines LSTM algorithm for real-time error compensation, while adaptive laser focal length adjustment ensures cutting accuracy. The "laser pre-marking + CNC mechanical cutting" double process mode is adopted to optimize cutting quality, and the DQN reinforcement learning algorithm is used to automatically feedback errors, realizing continuous optimization of parameters. The overall scheme improves cutting precision and efficiency through intelligent control, and reduces material waste.

[0043] In S2, the visible light laser uses a semiconductor laser source with a wavelength of 400-450 nm and a power range of 5-10 mW, which is used to clearly capture the detailed information of the wood surface, including texture and color details. The infrared laser uses an infrared laser source with a wavelength of 800-940 nm and a power range of 10-20 mW, which is used to deal with the situation where the wood surface color is deep or the reflectivity is low. The wavelength and power of the laser are adjusted in real time by the built-in photoelectric adjustment module according to the feedback information from the wood surface, which includes reflected light intensity and color, within 45-50 ms.

[0044] Specifically, the visible light laser uses a semiconductor laser source with a wavelength of 400-450 nm and a power range of 5-10 mW, which has a high color resolution and can clearly capture the detailed information of the wood surface, including wood texture, color changes, and surface structure characteristics, allowing the system to accurately identify the characteristics of the wood surface and provide more accurate path planning data in the subsequent cutting process. The infrared laser uses an infrared laser source with a wavelength of 800-940 nm and a power range of 10-20 mW, which is mainly used to adapt to the situation where the wood surface color is deep or the reflectivity is low. Due to the longer wavelength of infrared light, it has better penetration and can reduce measurement errors caused by differences in surface reflectivity, improving the stability of distance measurement. During distance measurement, the system adjusts the wavelength and power of the laser in real time based on the feedback information from the wood surface through the built-in high-precision photoelectric adjustment module, to adapt to different wood characteristics and environmental conditions. This photoelectric adjustment module can receive feedback data from the sensor in real time, including the reflected light intensity and color information of the wood surface, and dynamically adjust the wavelength and power of the laser within 45-50 ms based on these data, ensuring that the distance measurement accuracy always maintains within the predetermined range, avoiding measurement errors caused by changes in wood surface characteristics. In addition, the system also has adaptive adjustment function, which can automatically optimize the laser parameters in complex wood environment, making the measurement data more stable and reliable, thereby improving the precision and quality of subsequent cutting process. Through this dual-spectrum laser fusion distance measurement technology, the invention can effectively solve the problem of large distance measurement error of traditional single-wavelength laser distance measurement method when the wood surface characteristics change greatly, improve the adaptability to different types of wood, and provide higher precision positioning information for subsequent intelligent cutting process.

[0045] S3, features include texture, knot, crack and density unevenness, the optimized cutting path is used to avoid the defect area to avoid waste of wood, the AI algorithm adopts a U-Net neural network algorithm based on deep learning, the neural network structure includes 5 convolution layers, the convolution kernel sizes are 3*3, 5*5 and 7*7 respectively, the step is 1, and 3 fully connected layers, the number of neurons is 128, 64 and 32 respectively, which are used to analyze the features of the texture, knot, crack and density unevenness of the wood.

[0046] Specifically, first, the image data of the wood surface is obtained through a high-precision imaging system, and the internal density distribution of the wood is detected in combination with X-ray or laser scanning data. Then, the U-Net neural network processes these data, analyzes the texture direction, knot position, crack range and density change of the wood, and forms a complete wood feature map. For texture analysis, the system extracts the feature parameters of the wood surface gray level co-occurrence matrix, such as contrast, correlation, energy and entropy, to determine the texture direction and optimize the cutting direction; for knots, the target detection algorithm is used to identify the area, gray value and edge features of the circular or elliptical region in the image, to accurately determine the position and size of the knot and prevent the quality of the finished product from being reduced due to the cutting of the knife on the knot; for cracks, the length and depth of the crack are detected through the pixel gradient change and continuity analysis algorithm, and an avoidance path is generated; for the case of uneven density, the X-ray density imaging algorithm is used to identify high-density and low-density areas, and the cutting path is adjusted to optimize the performance of the finished product;

[0047] Based on the above analysis results, the core goal of optimizing the cutting path is to maximize the avoidance of defect areas while ensuring the maximum utilization of wood. The intelligent path optimization algorithm generates a cutting path according to the characteristics of the wood to minimize the cutting length, reduce waste, and avoid the impact of knots and cracks on the quality of the finished product by intelligently allocating cutting areas. The optimization algorithm can complete path adjustment within 100ms and update cutting sequence and process parameters in real time to adapt to different actual situations of wood. Through the AI intelligent path optimization technology of the present application, the utilization rate of wood can be effectively improved by 12%-15%, the processing loss is reduced, the production efficiency is improved, and the stability and consistency of the finished product quality are ensured, thereby greatly improving the precision and intelligent level of the cutting process of woodworking machinery.

[0048] In S4, the IMU inertial sensor is used to monitor the acceleration and angular velocity changes of the device in X, Y, Z three directions in real time, the acceleration measurement range is ±16g, the accuracy is ±0.01g, the angular velocity measurement range is ±2000dps, the accuracy is ±0.1dps, the deep learning algorithm adopts the long short-term memory network LSTM algorithm, which is used to analyze the data collected by the IMU sensor and the positioning data fed back by the laser positioning system, and to construct an error prediction model, and the error prediction model dynamically compensates the positioning error in 140-150ms, so that the cutting path accuracy is improved to within ±0.08mm.

[0049] Specifically, the long short-term memory network (LSTM) deep learning algorithm is used to analyze the motion data collected by the IMU sensor and the actual positioning data fed back by the laser positioning system, and to construct an error prediction model. The LSTM model stores historical data and dynamically updates short-term data, so that the system can predict the error change trend in different operating environments and complete error compensation calculation within 150ms. Specifically, the error prediction model first calculates the real-time displacement error using sensor data, analyzes the system error pattern in combination with historical data, and then generates an error correction value. Then, the system adjusts the laser cutting path and focal length to dynamically correct errors caused by device offset, vibration or environmental factors, ensuring that the cutting path is consistent with the theoretical path;

[0050] In addition, to further optimize the compensation accuracy, the system combines laser ranging data and uses deep fusion algorithm to synchronize sensor error correction, so that the final positioning error is controlled within ±0.08mm. This compensation mechanism not only applies to static precise positioning, but also adapts to dynamic cutting scenarios, improving the system's adaptability to complex working conditions. Through the laser error dynamic compensation technology of the present application, cutting deviation caused by device jitter or external interference can be effectively reduced, cutting precision can be improved, cutting quality consistency can be ensured, waste rate can be reduced, and processing efficiency can be improved, providing stable and reliable technical support for high-precision woodworking machinery cutting.

[0051] In S5, double-process cutting includes laser pre-scribing and mechanical cutting, the laser power of low-power laser pre-scribing is 3-5mW, the laser beam diameter is 0.1-0.2mm, the pre-scribing speed is 10-15mm / s, the cutting speed of CNC mechanical cutting equipment is 50-150mm / s, and the cutting accuracy is ±0.03mm.

[0052] Specifically, first, before cutting, low-power laser is used for pre-scribing to clearly define the cutting path and ensure the accuracy of the cutting process. This pre-scribing not only provides accurate reference for subsequent mechanical cutting, but also reduces cutting deviation caused by changes in wood density or different texture directions, improving cutting consistency and accuracy;

[0053] After the pre-scribing is completed, the CNC mechanical cutting equipment performs the final cutting, by combining the laser pre-scribing with the mechanical cutting, the process overcomes the deviation problem caused by the changes in wood hardness, knots or cracks, etc. in single mechanical cutting, while effectively reducing the burr, burning or edge tearing phenomenon generated in the cutting process. In addition, the double-process cutting method can reduce the cutting resistance of the mechanical cutter, thereby reducing the cutter wear, improving the service life of the equipment, and ensuring the smoothness of the cutting surface, significantly improving the wood processing accuracy, ensuring high-quality cutting, while reducing wood waste and improving production efficiency. Compared with the traditional single cutting process, the present application can improve the accuracy and reduce the waste caused by cutting errors through the double-process cooperative cutting, so that the final product quality is more stable and reliable.

[0054] In S6, the gas pressure range of the high-pressure gas flow jetting device is 0.5-0.8 MPa, the gas flow is 2-3 m 3 / min, the nozzle type is SV-10, the nozzle diameter is 2-3 mm, the water mist particle diameter of the water mist jetting device is 10-20 μm, the water spraying pressure is 0.3-0.5 MPa, the water spraying flow is 1-2 L / min, and the nozzle type is WS-20, the nozzle hole number is 10-15.

[0055] Specifically, through the strong power of high-pressure gas flow, the heat generated during cutting can be quickly and effectively removed, reducing the temperature of the cutting area, thereby preventing the wood from deforming, cracking or burning due to excessive temperature, ensuring the flatness and quality of the wood cutting surface, and the water mist jetting device further improves the cooling effect by spraying fine water mist, the water mist particle diameter is 10-20 μm, the water spraying pressure is set to 0.3-0.5 MPa, and the water spraying flow is 1-2 L / min. The nozzle type is WS-20, the nozzle hole number is 10-15, and a uniform water mist can be formed in the cutting area. The fine water mist not only helps to reduce the temperature of the cutting area and reduce the thermal stress on the wood surface, but also effectively prevents the wood surface from burning or appearing scorch marks, improving the quality and fineness of the wood surface after cutting;

[0056] The cooling and heat prevention system can provide stable cooling effect in the high-temperature environment of laser cutting, ensuring that the temperature fluctuation of the wood during cutting is minimized, thereby improving the cutting accuracy, reducing material waste, and effectively prolonging the service life of the cutting equipment. By using high-pressure gas flow and water mist in combination, the present application can provide a multi-dimensional cooling solution to meet the processing needs of different wood materials and ensure that the final cutting quality meets the high-precision requirements.

[0057] S7, the environmental parameters include temperature, humidity and dust concentration, the temperature monitoring accuracy is ±0.2℃, the humidity monitoring accuracy is ±2%RH, the dust concentration monitoring accuracy is ±1mg / m 3 , the negative pressure dust collection system, when the dust concentration is lower than 50mg / m 3 , the dust collection strength is adjusted to 200-300Pa, when the dust concentration is higher than 80mg / m 3 , the dust collection strength is adjusted to 400-500Pa.

[0058] Specifically, the intelligent environment adjustment system continuously monitors the working environment to ensure stable and reliable cutting process. The temperature control accuracy is ±0.2℃, the humidity detection error is not more than ±2%RH, and the dust concentration sensing accuracy is controlled within ±1mg / m 3 , the cutting parameters are accurately adjusted to reduce the environmental impact. The negative pressure dust removal module automatically adjusts the operating state according to the dust level, maintaining 200-300Pa at low value and increasing to 400-500Pa at high value, quickly removing particulate matter and ensuring stable light beam. This system optimizes the working environment, reduces interference factors, improves processing quality, and effectively prolongs the service life of the equipment.

[0059] S8, real-time detection of wood thickness change, using thickness sensor, detection accuracy is ±0.05mm, detection frequency is 450-500Hz, automatic adjustment of laser focal length uses electric zoom lens.

[0060] Specifically, the system real-time monitors the wood thickness change and automatically adjusts the laser focal length to ensure the laser always maintains the best focusing state, thereby improving the cutting accuracy and reducing the error caused by focal point deviation. The high-precision thickness sensor continuously detects the wood thickness at a frequency of 450-500Hz, with an accuracy of ±0.05mm, and transmits the data to the control system. After the system calculates the thickness change, it drives the electric zoom lens to automatically adjust the laser focal length to ensure the focal point always at the best cutting depth. This system can dynamically adapt to the thickness difference of wood, avoiding the phenomenon of cutting too deep or too shallow caused by thickness change, improving the processing consistency. At the same time, it reduces the need for repeated adjustments, improves cutting efficiency, reduces material waste and equipment wear, and ensures the stability of long-time operation.

[0061] S9, automatic feedback mechanism for real-time recording of error data in the cutting process, error data includes positioning error, cutting path deviation, cutting depth error, recording frequency is 10 times / cutting process, reinforcement learning algorithm uses deep Q network DQN algorithm, after completing 20 times of cutting, within 300-400ms, according to the recorded error data, the cutting parameters are optimized, the cutting error gradually decreases at a speed of 0.02mm per optimization, and the cutting efficiency gradually increases at a speed of 2% per optimization.

[0062] Specifically, the automatic feedback and optimization system improves cutting accuracy and efficiency through real-time error recording and reinforcement learning algorithms. The system records 10 error data points during each cutting process, including positioning errors, path deviations, and depth errors, and uses a deep Q-network (DQN) algorithm for optimization. After completing 20 cuts, the system analyzes error data within 300-400 ms and dynamically adjusts cutting parameters to gradually reduce cutting errors by 0.02 mm each optimization, while cutting efficiency improves by 2%. This mechanism ensures continuous self-improvement of the system, reducing accuracy deviations, improving processing consistency, and achieving efficient and precise intelligent cutting in the long run.

[0063] While embodiments of the present application have been shown and described with reference to a few embodiments, it will be understood by those skilled in the art that various changes, modifications, substitutions, and alterations can be made hereto without departing from the principles and spirit of the application, the scope of which is defined by the appended claims and their equivalents.

Claims

1. High precision cutting process for woodworking machines based on laser positioning, characterized in that, Comprise the following steps: S1, laser positioning technology application: through laser sensor scanning wood surface to obtain wood data, and then through the use of laser measurement to obtain preliminary positioning information; S2, dual-spectrum laser fusion ranging: using dual-spectrum laser fusion technology, combining visible light laser and infrared laser, by adjusting the wavelength and power of laser in real time, keeping the ranging accuracy; S3, AI intelligent path optimization: using AI algorithm to analyze the characteristics of wood, combining with the structure characteristics of wood to dynamically optimize the cutting path, before each cutting, according to the real-time scanning data of wood to adjust the cutting order and process; S4, laser error dynamic compensation: in the cutting process, using laser system to monitor the error caused by vibration, temperature and humidity change or equipment offset in real time, using IMU inertial sensor and deep learning algorithm to dynamically compensate the positioning error; S5, multi-process collaborative cutting: through the use of double-process cutting method, pre-marking is carried out by low-power laser marking during processing, which is used to define the cutting path, and then CNC mechanical cutting equipment is used for accurate cutting; S6, cooling and heat prevention technology: in the process of laser cutting, high-pressure airflow or water mist spraying device is equipped to reduce the temperature of cutting area and avoid the adverse effects of heat accumulation on wood surface; S7, intelligent environment adjustment: real-time monitoring of factory environment parameters to automatically adjust the working state of laser cutting system, through negative pressure dust collection system to automatically adjust the dust collection intensity according to the dust concentration to prevent dust from interfering with the laser positioning system; S8, adaptive laser focal length adjustment: using adaptive laser focal length adjustment mechanism, through real-time detection of wood thickness change, automatically adjusting the laser focal length, so that the laser cutting always maintains the best focal point, improving the cutting precision; S9, automatic feedback and optimization learning: using automatic feedback mechanism, real-time recording of error in each cutting process, through reinforcement learning algorithm to continuously optimize cutting parameters, self-learning the best cutting path and parameter adjustment strategy, improving cutting precision and efficiency with the increase of use time.

2. Laser positioning based high precision cutting process for woodworking machines according to claim 1, characterized in that: In S1, the wood data is used to obtain three-dimensional surface data of wood based on laser reflection principle to create digital model of wood, the measurement frequency of laser measurement is set to 2000 times / s, the positioning error is controlled to ±0.08mm, the laser sensor uses line laser sensor and plane laser sensor, the scanning frequency of line laser sensor is 1000-1200Hz, the measurement range is 0-80m, the accuracy is ±1mm, which is used to obtain linear profile data of wood surface, the scanning rate of plane laser sensor is 45000-50000 points / s, the measurement accuracy is ±2mm, which is used to obtain two-dimensional plane data of wood surface.

3. The high precision cutting process for laser positioning based woodworking machinery as claimed in claim 1 wherein: In S2, the visible light laser uses a semiconductor laser source with a wavelength of 400-450 nm and a power range of 5-10 mW to clearly capture the detailed information of the wood surface, including texture and color details. The infrared laser uses an infrared laser source with a wavelength of 800-940 nm and a power range of 10-20 mW to deal with situations where the wood surface is dark or has low reflectivity. The real-time adjustment of the laser wavelength and power is achieved through an internal photoelectric adjustment module based on feedback information from the wood surface, including reflected light intensity and color. The adjustment is completed within 45-50 ms.

4. The high precision cutting process for laser positioning based woodworking machinery as claimed in claim 1 wherein: In S3, the features include texture, knots, cracks, and uneven density. The optimized cutting path is designed to avoid defect areas to minimize wood waste. The AI algorithm uses a deep learning-based U-Net neural network algorithm with a neural network structure consisting of 5 convolutional layers with kernel sizes of 3x3, 5x5, and 7x7, a step size of 1, and 3 fully connected layers with neuron counts of 128, 64, and 32, respectively. This algorithm is used to analyze the features of wood texture, knots, cracks, and uneven density.

5. The high precision cutting process for laser positioning based woodworking machinery as claimed in claim 1 wherein: In S4, the IMU inertial sensor is used to monitor the acceleration and angular velocity changes of the device in X, Y, and Z directions in real time. The acceleration measurement range is ±16g with an accuracy of ±0.01g, and the angular velocity measurement range is ±2000dps with an accuracy of ±0.1dps. The deep learning algorithm uses a long short-term memory network (LSTM) algorithm to analyze the data collected by the IMU sensor and the positioning data feedback from the laser positioning system, and constructs an error prediction model. The error prediction model dynamically compensates for positioning errors within 140-150 ms, improving the cutting path accuracy to within ±0.08 mm.

6. The high precision cutting process for laser positioning based woodworking machinery as claimed in claim 1 wherein: In S5, the double-process cutting includes laser pre-marking and mechanical cutting. The low-power laser pre-marking has a laser power of 3-5 mW and a laser beam diameter of 0.1-0.2 mm, with a pre-marking speed of 10-15 mm / s. The CNC mechanical cutting equipment has a cutting speed of 50-150 mm / s and a cutting accuracy of ±0.03 mm.

7. The high precision cutting process for laser positioning based woodworking machinery as claimed in claim 1 wherein: The gas pressure range of the high-pressure gas jet device in S6 is 0.5-0.8 MPa, and the gas flow is 2-3 m 3 / min, the nozzle is of SV-10 type, the nozzle orifice diameter is 2-3 mm, the water mist particle diameter of the water mist jet device is 10-20 μm, the water spraying pressure is 0.3-0.5 MPa, the water spraying flow is 1-2 L / min, and the fine water mist is formed by a nozzle of WS-20 type, and the nozzle orifice number is 10-15.

8. The high precision cutting process for laser positioning based woodworking machinery as claimed in claim 1 wherein: In S7, the environmental parameters include temperature, humidity and dust concentration, the accuracy of temperature monitoring is ±0.2℃, the accuracy of humidity monitoring is ±2%RH, and the accuracy of dust concentration monitoring is ±1mg / m 3 The negative pressure dust collection system adjusts the dust collection strength to 200-300Pa when the dust concentration is lower than 50mg / m 3 The negative pressure dust collection system adjusts the dust collection strength to 400-500Pa when the dust concentration is higher than 80mg / m 3 The negative pressure dust collection system adjusts the dust collection strength to 400-500Pa when the dust concentration is higher than 80mg / m 9. The high precision cutting process for laser positioning based woodworking machinery as claimed in claim 1 wherein: In S8, the real-time detection of wood thickness changes uses a thickness sensor with a detection accuracy of ±0.05 mm and a detection frequency of 450-500 Hz. The automatic adjustment of the laser focal length uses an electric zoom lens.

10. The high precision cutting process for laser positioning based woodworking machinery as claimed in claim 1 wherein: In S9, the automatic feedback mechanism is used to record error data during the cutting process in real time. The error data includes positioning errors, cutting path deviations, and cutting depth errors, with a recording frequency of 10 times per cutting process. The reinforcement learning algorithm uses a deep Q-network (DQN) algorithm to optimize cutting parameters based on recorded error data within 300-400 ms after completing 20 cuts. The cutting error gradually decreases at a rate of 0.02 mm per optimization, and the cutting efficiency gradually increases at a rate of 2% per optimization.

Images