Digital welder energy-saving mode switching method and system

Abstract

The present application relates to energy-saving mode switching method technical field, especially to digital welding machine energy-saving mode switching method and system, the method comprises the following steps: acquire reflected millimeter wave signal data; the reflected millimeter wave signal data is handled to Doppler frequency shift, and dynamic target signal data is obtained;Dynamic target signal data is handled to phase demodulation, and complex amplitude signal is obtained;Complex amplitude signal is handled to phase estimation, and hand phase change information data is obtained;Hand phase change information data is handled to feature conversion, and operation gesture data is obtained;Operation gesture data is used to operate hand posture reconstruction, and operation hand model is obtained;The present application obtains hand related information by non-contact mode, and combines hand posture, grip ability and machine interactive force data, provides more comprehensive basis for the switching of digital welding machine energy-saving mode, realizes more intelligent and efficient energy consumption management.

Description

Technical Field

[0001] The present invention relates to the technical field of energy-saving mode switching methods, and in particular to an energy-saving mode switching method and system for a digital welding machine. Background Art

[0002] The welding industry is a critical component of the manufacturing industry and is widely used in various industrial sectors, such as automotive, aerospace, power equipment, and building structures. Welding is an energy-intensive process, and welding machines consume significant amounts of energy when operating at high power. The advent of digital welders can significantly improve energy conversion efficiency and reduce energy loss. By employing energy-saving mode switching methods and systems, digital welders can reduce power consumption when the welder is not in use or under light load, reducing energy consumption. These methods and systems automatically adjust to the welding process requirements, ensuring the welder maintains optimal operating conditions under varying loads and process conditions. However, current digital welders on the market may experience a delay when switching to energy-saving mode. This means there may be a lag between switching from high-power mode to energy-saving mode, resulting in wasted energy and reduced efficiency. Furthermore, some digital welders lack a user-friendly energy-saving mode switching interface or functionality, which can cause confusion and inconvenience when operating and adjusting energy-saving modes, impacting the user experience. Summary of the Invention

[0003] Based on this, it is necessary to provide a method and system for switching energy-saving modes of a digital welding machine to solve at least one of the above technical problems.

[0004] To achieve the above object, a method for switching energy-saving mode of a digital welding machine includes the following steps:

[0005] Step S1: obtaining millimeter wave signal data reflected by a digital welder; performing Doppler frequency shift processing on the millimeter wave signal data reflected by the digital welder to obtain dynamic hand target signal data;

[0006] Step S2: performing phase demodulation processing on the dynamic hand target signal data to obtain a complex amplitude signal; performing phase estimation processing on the complex amplitude signal to obtain hand phase change information data;

[0007] Step S3: performing feature conversion processing on the hand phase change information data to obtain operation gesture data; reconstructing the operation hand posture using the operation gesture data to obtain an operation hand model; and performing model segmentation on the operation hand model to obtain the hand-machine interaction area and part geometry;

[0008] Step S4: performing morphological analysis on the geometrical morphology of the part to obtain part structural characteristic data; performing grasping ability evaluation based on the part structural characteristic data to obtain hand grasping ability data;

[0009] Step S5: Dynamically analyzing the hand-machine interaction area to obtain hand deformation information data; performing an inversion operation on the hand deformation information data based on a preset hand force and deformation relationship to obtain hand applied force data;

[0010] Step S6: performing interaction force analysis on the hand gripping ability data and the hand applied force data to obtain hand-machine interaction force data; and switching the energy-saving mode of the digital welding machine according to the hand-machine interaction force data.

[0011] The present invention uses a millimeter wave sensor or radar equipment to obtain millimeter wave signal data reflected by the target object, performs Doppler frequency shift processing on the reflected millimeter wave signal, obtains dynamic information of the target object, including speed and movement direction, performs phase demodulation processing on the dynamic information, obtains complex amplitude signal and hand phase change information data, performs phase estimation processing on the complex amplitude signal, extracts operation gesture data, uses the operation gesture data to reconstruct the hand posture, restores the model of the operation hand, cuts the hand model, obtains the geometric shape of the hand-machine interaction area and part, and calculates the part geometry. Morphological analysis of the morphology can extract part structure feature data, and grasping ability assessment is performed based on the part structure feature data to obtain hand grasping ability data. Dynamic analysis of the hand-machine interaction area can obtain hand deformation information data. Based on the preset hand force and deformation relationship, an inversion operation can be performed to obtain hand applied force data. Interaction force analysis of the hand grasping ability data and the hand applied force data can be performed to obtain hand-machine interaction force data. Based on the hand-machine interaction force data, the digital welder is switched to an energy-saving mode to achieve more efficient welding operations. Therefore, the present invention obtains hand-related information in a non-contact manner, and combines hand posture, grasping ability and machine interaction force data to provide a more comprehensive basis for switching the energy-saving mode of the digital welder, thereby achieving more intelligent and efficient energy consumption management.

[0012] The present invention utilizes reflected millimeter wave signals for data acquisition, eliminating the need for direct hand contact and avoiding the limitations of contact measurement with traditional sensors. Through phase demodulation and estimation, the dynamic target signal is converted into a complex amplitude signal. Hand phase change information is extracted and feature conversion is performed, enabling recognition of hand gestures and reconstruction of hand posture. Hand models are then segmented and subjected to morphological analysis of part geometry to obtain part structural feature data, further enabling assessment of the hand's grasping ability. This helps understand the hand's function and adaptability to different operations. Dynamic analysis of the hand-machine interaction area and deformation information data allows estimation of the force applied by the hand. Based on a preset hand force-deformation relationship, an inversion operation is performed to obtain hand-applied force data, providing a more comprehensive understanding of the force interaction between the hand and the object. Combined with the hand's grasping ability data and hand-applied force data, interaction force analysis is performed to obtain more comprehensive hand-machine interaction force data. This helps assess the hand's needs and adaptability to digital welding machine operations. Using this hand-machine interaction force data, the digital welding machine's energy-saving mode can be switched according to pre-defined rules or algorithms. This allows the energy consumption of the digital welder to be adjusted according to the actual needs and capabilities of the hand, improving energy efficiency. Therefore, the present invention obtains hand-related information in a contactless manner and combines it with hand posture, gripping ability, and machine interaction force data to provide a more comprehensive basis for switching the digital welder's energy-saving mode, achieving more intelligent and efficient energy management.

[0013] Preferably, step S1 includes:

[0014] Step S11: using a millimeter wave sensor to collect data from the digital welder to obtain millimeter wave signal data reflected by the digital welder;

[0015] Step S12: using a window function to perform weighted processing on the digital welding machine reflected millimeter wave signal data to obtain weighted reflected millimeter wave signal data;

[0016] Step S13: performing signal segmentation on the weighted reflected millimeter wave signal data to obtain a segmented signal data set;

[0017] Step S14: performing Fourier transform on the segmented signal data set to obtain a frequency domain signal data set;

[0018] Step S15: performing frequency domain analysis on the frequency domain signal data set to obtain frequency signal change data;

[0019] Step S16: performing Doppler frequency shift calculation on the frequency signal change data to obtain dynamic hand target signal data.

[0020] Preferably, step S2 includes the following steps:

[0021] Step S21: performing bandpass filtering on the dynamic hand target signal data to obtain prominent phase signal data;

[0022] Step S22: performing phase demodulation processing on the prominent phase signal data to obtain a complex amplitude signal;

[0023] Step S23: performing a differential operation on the complex amplitude signal to obtain a differential signal;

[0024] Step S24: Calculate the phase change according to the differential signal to obtain hand phase change information data.

[0025] Preferably, step S3 includes the following steps:

[0026] Step S31: using a phase gesture conversion algorithm to perform feature conversion on the hand phase change information data to obtain effective gesture feature data;

[0027] Step S32: performing gesture recognition processing on the valid gesture feature data using a machine learning algorithm to obtain operation gesture data;

[0028] Step S33: constructing a three-dimensional model based on the operation gesture data to obtain a reconstructed hand posture model;

[0029] Step S34: geometrically subdividing the reconstructed hand posture model to obtain an operation gesture model;

[0030] Step S35: cutting the operating hand model to obtain the hand-machine interaction area and part geometry.

[0031] Preferably, the phase gesture conversion algorithm in step S31 is as follows:

[0032]

[0033] Among them, Y(t) represents the operation gesture data sequence, that is, the gesture features after the final conversion, X(t) represents the hand phase change information data sequence, that is, the input original gesture data, Δt represents the small time difference, which is the time interval used for the derivative function. is a derivative operator used to calculate the derivative of a function with respect to time. X(t+Δt) represents the hand phase change information data within a small time interval Δt after time t, where t represents time. It represents the angle feature of normalizing the hand phase change information data to [-1, 1]. π represents pi, which is used to convert the angle into radians.

[0034] Preferably, step S4 includes the following steps:

[0035] Step S41: extracting geometric features of the part geometry based on morphological analysis to obtain part structural feature data;

[0036] Step S42: performing mechanical structure analysis on the part structural characteristic data to obtain part mechanical performance data;

[0037] Step S43: performing specific feature detection on the part structure feature data to obtain various hand feature data, wherein the specific feature detection includes measuring finger length and palm width, observing muscle tissue and joint flexibility, and examining the hand bone structure;

[0038] Step S44: performing a grasping ability assessment based on the part mechanical performance data and the hand characteristic data to obtain the hand grasping ability data.

[0039] Preferably, step S5 includes the following steps:

[0040] Step S51: Dynamically analyze the hand-machine interaction area to obtain action sequence data of the interaction area;

[0041] Step S52: performing action sequence comparison on the interactive area action sequence data to obtain comparative analysis data;

[0042] Step S53: extracting hand deformation information data from the comparative analysis data;

[0043] Step S54: Based on the preset hand force and deformation relationship, an inversion operation is performed on the hand deformation information data using a force and deformation inversion algorithm to obtain the hand applied force data.

[0044] Preferably, the force and deformation inversion algorithm in step S54 is as follows:

[0045]

[0046] Among them, F(x,a) is the force exerted by the hand at position x at time t, that is, the result obtained by inversion, and u(x,s) is the deformation information at position x, including displacement, deformation and other information. is a gradient operator used to calculate the gradient of the deformation field u(x,s), Ω is the overall deformation area of ​​the hand, and defines the distribution of the deformation information u(x,s) within the entire hand range, e -a is an exponential term, which represents time attenuation, indicating that the effect of force gradually decreases with the passage of time. cos(a) is a trigonometric function term, which represents the periodic force change at time a. T is the upper limit of time integration. is the partial derivative, is the partial derivative of the deformation function u with respect to the spatial coordinate s, is a weight or influence term based on the distance between points x and s, and ds is an infinitesimal element in the spatial domain Ω.

[0047] Preferably, step S6 includes the following steps:

[0048] Step S61: performing interaction force analysis based on the hand grasping ability data and the hand applied force data to obtain hand-machine interaction force data;

[0049] Step S62: Perform interaction mechanics analysis on the hand-machine interaction force data to obtain machine mastery data;

[0050] Step S63: performing interval positioning on the machine mastery data according to the preset machine mastery data interval to obtain the energy-saving mode interval;

[0051] Step S64: switching the energy-saving mode of the digital welding machine according to the energy-saving mode range.

[0052] The present invention also provides a digital welding machine energy-saving mode switching system for executing the above-mentioned digital welding machine energy-saving mode switching method, comprising:

[0053] The hand acquisition module is used to obtain the millimeter wave signal data reflected by the digital welder; perform Doppler frequency shift processing on the millimeter wave signal data reflected by the digital welder to obtain dynamic hand target signal data;

[0054] The data processing module is used to perform phase demodulation processing on the dynamic hand target signal data to obtain a complex amplitude signal; and perform phase estimation processing on the complex amplitude signal to obtain hand phase change information data;

[0055] The model reconstruction module is used to perform feature conversion processing on the hand phase change information data to obtain the operation gesture data; use the operation gesture data to reconstruct the operation hand posture to obtain the operation hand model; and perform model cutting on the operation hand model to obtain the hand-machine interaction area and part geometry;

[0056] The ability assessment module is used to perform morphological analysis on the geometric shape of the part to obtain the structural feature data of the part; and to perform grasping ability assessment based on the structural feature data of the part to obtain the hand grasping ability data;

[0057] The deformation inversion module is used to perform dynamic analysis on the hand-machine interaction area to obtain hand deformation information data; based on the preset hand force and deformation relationship, the hand deformation information data is inverted to obtain the hand applied force data;

[0058] The mode switching module is used to perform interactive force analysis on the hand grasping ability data and the hand applied force data to obtain the hand-machine interactive force data; and switch the energy-saving mode of the digital welding machine according to the hand-machine interactive force data.

[0059] The present invention uses millimeter wave sensors to collect data and perform Doppler frequency shift processing, phase demodulation processing, feature conversion processing and other steps to evaluate hand grasping ability and perform hand-machine interaction force analysis, thereby realizing the switching of the digital welding machine's energy-saving mode.

[0060] The present invention utilizes reflected millimeter wave signals for data acquisition, eliminating the need for direct hand contact and avoiding the limitations of contact measurement with traditional sensors. Through phase demodulation and estimation, the dynamic target signal is converted into a complex amplitude signal. Hand phase change information is extracted and feature conversion is performed, enabling recognition of hand gestures and reconstruction of hand posture. Hand models are then segmented and subjected to morphological analysis of part geometry to obtain part structural feature data, further enabling assessment of the hand's grasping ability. This helps understand the hand's function and adaptability to different operations. Dynamic analysis of the hand-machine interaction area and deformation information data allows estimation of the force applied by the hand. Based on a preset hand force-deformation relationship, an inversion operation is performed to obtain hand-applied force data, providing a more comprehensive understanding of the force interaction between the hand and the object. Combined with the hand's grasping ability data and hand-applied force data, interaction force analysis is performed to obtain more comprehensive hand-machine interaction force data. This helps assess the hand's needs and adaptability to digital welding machine operations. Using this hand-machine interaction force data, the digital welding machine's energy-saving mode can be switched according to pre-defined rules or algorithms. In this way, the energy consumption of the digital welding machine can be adjusted according to the actual needs and capabilities of the hand, thereby improving energy utilization efficiency. The present invention can obtain hand-related information in a non-contact manner, and combine hand posture, grasping ability and machine interaction force data to provide a more comprehensive basis for switching the energy-saving mode of the digital welding machine, thereby achieving more intelligent and efficient energy consumption management. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Figure 1 A schematic flow chart of the steps of a method for switching energy-saving modes of a digital welding machine according to an embodiment;

[0062] Figure 2 for Figure 1 Detailed implementation steps of step S3 in FIG.

[0063] Figure 3 for Figure 1 Detailed implementation steps of step S4 in FIG.

[0064] Figure 4 for Figure 1 Detailed implementation steps of step S5;

[0065] The purpose, features and advantages of the present invention will be further described with reference to the accompanying drawings and in conjunction with the embodiments. DETAILED DESCRIPTION

[0066] The following is a clear and complete description of the technical method of the present invention in conjunction with the accompanying drawings. Obviously, the embodiments described are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without making any creative work are within the scope of protection of the present invention.

[0067] In addition, the accompanying drawings are merely schematic illustrations of the present invention and are not necessarily drawn to scale. Identical reference numerals in the figures denote identical or similar parts, and thus repetitive descriptions thereof will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities that do not necessarily correspond to physically or logically separate entities. These functional entities may be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor and / or microcontroller approaches.

[0068] It should be understood that although the terms "first," "second," and the like may be used herein to describe various elements, these elements should not be limited by these terms. These terms are used solely to distinguish one element from another. For example, a first element may be referred to as a second element, and similarly, a second element may be referred to as a first element, without departing from the scope of the exemplary embodiments. The term "and / or" as used herein includes any and all combinations of one or more of the listed associated items.

[0069] To achieve this, please refer to Figures 1 to 4 The present invention provides a method for switching energy-saving modes of a digital welding machine, comprising the following steps:

[0070] Step S1: obtaining millimeter wave signal data reflected by a digital welder; performing Doppler frequency shift processing on the millimeter wave signal data reflected by the digital welder to obtain dynamic hand target signal data;

[0071] Step S2: performing phase demodulation processing on the dynamic hand target signal data to obtain a complex amplitude signal; performing phase estimation processing on the complex amplitude signal to obtain hand phase change information data;

[0072] Step S3: performing feature conversion processing on the hand phase change information data to obtain operation gesture data; reconstructing the operation hand posture using the operation gesture data to obtain an operation hand model; and performing model segmentation on the operation hand model to obtain the hand-machine interaction area and part geometry;

[0073] Step S4: performing morphological analysis on the geometrical morphology of the part to obtain part structural characteristic data; performing grasping ability evaluation based on the part structural characteristic data to obtain hand grasping ability data;

[0074] Step S5: Dynamically analyzing the hand-machine interaction area to obtain hand deformation information data; performing an inversion operation on the hand deformation information data based on a preset hand force and deformation relationship to obtain hand applied force data;

[0075] Step S6: performing interaction force analysis on the hand gripping ability data and the hand applied force data to obtain hand-machine interaction force data; and switching the energy-saving mode of the digital welding machine according to the hand-machine interaction force data.

[0076] The present invention uses a millimeter wave sensor or radar equipment to obtain millimeter wave signal data reflected by the target object, performs Doppler frequency shift processing on the reflected millimeter wave signal, obtains dynamic information of the target object, including speed and movement direction, performs phase demodulation processing on the dynamic information, obtains complex amplitude signal and hand phase change information data, performs phase estimation processing on the complex amplitude signal, extracts operation gesture data, uses the operation gesture data to reconstruct the hand posture, restores the model of the operation hand, cuts the hand model, obtains the geometric shape of the hand-machine interaction area and part, and calculates the part geometry. Morphological analysis of the morphology can extract part structure feature data, and grasping ability assessment is performed based on the part structure feature data to obtain hand grasping ability data. Dynamic analysis of the hand-machine interaction area can obtain hand deformation information data. Based on the preset hand force and deformation relationship, an inversion operation can be performed to obtain hand applied force data. Interaction force analysis of the hand grasping ability data and the hand applied force data can be performed to obtain hand-machine interaction force data. Based on the hand-machine interaction force data, the digital welder is switched to an energy-saving mode to achieve more efficient welding operations. Therefore, the present invention obtains hand-related information in a non-contact manner, and combines hand posture, grasping ability and machine interaction force data to provide a more comprehensive basis for switching the energy-saving mode of the digital welder, thereby achieving more intelligent and efficient energy consumption management.

[0077] In the embodiment of the present invention, reference Figure 1 The above is a schematic flow chart of the steps of the method for switching the energy-saving mode of a digital welding machine according to the present invention. In this example, the method for switching the energy-saving mode of a digital welding machine includes the following steps:

[0078] Step S1: obtaining millimeter wave signal data reflected by a digital welder; performing Doppler frequency shift processing on the millimeter wave signal data reflected by the digital welder to obtain dynamic hand target signal data;

[0079] In an embodiment of the present invention, a suitable millimeter wave radar system is provided, including components such as an antenna, a radio frequency front end, a transmitter, and a receiver. Appropriate antenna configuration and beamforming technology are used to focus the transmitted millimeter wave signal and receive the reflected signal, transmit the millimeter wave signal, receive the reflected signal, and convert the received signal into digital data. This can be accomplished through an analog-to-digital converter (ADC), using digital signal processing (DSP) technology to process the collected millimeter wave signal data, applying a Doppler shift algorithm, and detecting the movement of the target by comparing the signal frequency changes within consecutive time intervals. This usually includes Fourier transform or other frequency domain analysis methods to extract Doppler shift information from the processed signal, which reflects the speed of the target, and converting the extracted Doppler shift information into dynamic target signal data, which can be a series of time-related speed information.

[0080] Step S2: performing phase demodulation processing on the dynamic hand target signal data to obtain a complex amplitude signal; performing phase estimation processing on the complex amplitude signal to obtain hand phase change information data;

[0081] In an embodiment of the present invention, a complex amplitude signal is extracted from dynamic target signal data. This can be achieved by using in-phase and quadrature-phase components, using methods such as Hilbert transform, to convert the complex amplitude signal into phase information. This typically involves using an arctan function or other phase demodulation algorithm to process the phase sequence to extract information about the phase change over time. The phase difference between adjacent moments can be calculated, i.e., the rate of change of the phase can be estimated. This can be achieved by simply calculating the phase difference between adjacent time points, or by applying a more complex differential algorithm, using Hilbert transform or other methods to extract the complex amplitude signal, performing phase demodulation on the complex amplitude signal to obtain phase information, and extracting the hand phase change information from the phase information by calculating the phase difference or other related methods.

[0082] Step S3: performing feature conversion processing on the hand phase change information data to obtain operation gesture data; reconstructing the operation hand posture using the operation gesture data to obtain an operation hand model; and performing model segmentation on the operation hand model to obtain the hand-machine interaction area and part geometry;

[0083] In an embodiment of the present invention, features, such as frequency domain features and time domain features, are extracted from hand phase change information data. This may include using Fourier transform, wavelet transform, or other feature extraction techniques to perform dimensionality reduction processing on the extracted features to reduce data complexity and redundant information. Common methods include principal component analysis (PCA) or other dimensionality reduction techniques, selecting an appropriate hand posture model, such as a joint angle-based model or a neural network-based model, training or calibrating the selected hand posture model using gesture data to establish a mapping relationship between hand phase change information and hand posture, using the trained model to map the hand phase change information to a specific hand posture to achieve hand posture reconstruction, and using the hand posture model to segment the entire hand model to distinguish between parts such as the palm and fingers, determining areas in the hand model related to machine interaction, such as the palm or specific finger areas, and establishing geometric morphological models of various parts of the hand based on the segmented hand model. This may include information such as finger length and palm shape.

[0084] Step S4: performing morphological analysis on the geometrical morphology of the part to obtain part structural characteristic data; performing grasping ability evaluation based on the part structural characteristic data to obtain hand grasping ability data;

[0085] In embodiments of the present invention, to obtain an image or three-dimensional model of the geometric morphology of a hand part, a camera, depth sensor, or other device may be used to pre-process the image, including denoising and edge detection, to better extract the structural information of the part. Morphological operations, such as erosion, dilation, opening, and closing, are then applied to highlight the structural features of the part and remove unnecessary details. Key structural features, such as convexity and concavity, curvature, length, and width, are then extracted from the morphologically processed image or model. The extracted structural features are then quantified for subsequent analysis and evaluation. This may involve feature measurement and statistics, and indicators for grasping ability assessment are established based on the previously extracted structural features of the part. These indicators may include finger flexibility, palm stability, strength of various parts, etc. Machine learning models or rule-based methods can be used to establish a mapping model from part structural features to grasping ability. If a machine learning model is used, it is necessary to use known grasping ability data for training or calibration so that the model can accurately evaluate grasping ability. The part structural feature data is input into the established model to obtain an estimate of the hand grasping ability. The model output is interpreted and converted into understandable hand grasping ability data, such as strength level, stability score, etc.

[0086] Step S5: Dynamically analyzing the hand-machine interaction area to obtain hand deformation information data; performing an inversion operation on the hand deformation information data based on a preset hand force and deformation relationship to obtain hand applied force data;

[0087] In an embodiment of the present invention, data is analyzed to identify and extract information related to hand deformation. This may include changes in finger joints, deformation of the palm area, etc. Key deformation features, such as deformation degree, position, speed, time series, etc., are extracted from the collected data. Through experiments or previous studies, the relationship between hand deformation and force is obtained. This may require calibration experiments to measure the applied force under different deformation conditions. Based on experimental data or theoretical assumptions, a mathematical or statistical model between hand deformation and applied force is established. The established hand force and deformation relationship model is used to convert the hand deformation information data obtained from the sensor into corresponding applied force data. The converted force data is processed, which may include operations such as correction, smoothing, and unit conversion. The data is then interpreted and converted into a form that is easy to understand and apply, such as information such as the magnitude, direction, and point of application of the force.

[0088] Step S6: performing interaction force analysis on the hand gripping ability data and the hand applied force data to obtain hand-machine interaction force data; switching the energy-saving mode of the digital welding machine according to the hand-machine interaction force data;

[0089] In an embodiment of the present invention, the method or apparatus established in step S5 is used to obtain data on the force applied by the hand in the machine interaction area. This is combined with the hand gripping ability data and applied force data to perform interaction force analysis. This may require pairing, time synchronization, or correlation analysis of these two types of data to understand the relationship between gripping force and applied force. Determining interaction force data: Based on the analysis results, actual force data for the hand-machine interaction area is obtained. This data reflects the force applied by the hand on the machine and the characteristics of the grip. Establishing an energy-saving mode strategy: Based on the analysis results of the interaction force data, a strategy for switching to an energy-saving mode for the digital welder is formulated. This may involve setting force thresholds or monitoring force change trends. System integration and control: The designed strategy is integrated with the digital welder's control system to ensure real-time monitoring of hand interaction force data and automatically switch the welder's energy-saving mode according to preset conditions. This ensures that the system can continuously monitor hand interaction force data and make adjustments based on actual conditions. This may require optimizing the energy-saving mode thresholds or switching conditions to accommodate changes in hand interaction force under different working conditions.

[0090] Preferably, step S1 includes the following steps:

[0091] Step S11: using a millimeter wave sensor to collect data from the digital welder to obtain millimeter wave signal data reflected by the digital welder;

[0092] Step S12: using a window function to perform weighted processing on the digital welding machine reflected millimeter wave signal data to obtain weighted reflected millimeter wave signal data;

[0093] Step S13: performing signal segmentation on the weighted reflected millimeter wave signal data to obtain a segmented signal data set;

[0094] Step S14: performing Fourier transform on the segmented signal data set to obtain a frequency domain signal data set;

[0095] Step S15: performing frequency domain analysis on the frequency domain signal data set to obtain frequency signal change data;

[0096] Step S16: performing Doppler frequency shift calculation on the frequency signal change data to obtain dynamic hand target signal data.

[0097] The present invention uses millimeter wave sensors for data acquisition, capturing millimeter wave reflection signals from target objects in the environment. Millimeter waves can penetrate some obstacles, making them highly adaptable to diverse environments. Using a window function to weight the reflected millimeter wave signals can improve signal processing accuracy and anti-interference capabilities. The window function can effectively reduce problems such as spectral leakage and enhance the performance of subsequent signal processing. Segmenting the weighted reflected millimeter wave signals into distinct time domain segments facilitates better analysis of target motion and characteristics. This helps address multi-target scenarios and dynamic environments. Fourier transforming the segmented signal dataset converts the signal from the time domain to the frequency domain. This facilitates analysis of the signal's frequency components in the frequency domain, clarifying the target's characteristics. Frequency domain analysis helps identify target frequency information, such as its motion state or vibration frequency. This provides more contextual information about the target, and Doppler shift calculations can be used to determine its velocity. This is critical for target tracking and detection in dynamic scenarios, particularly in applications requiring real-time feedback.

[0098] In embodiments of the present invention, millimeter wave sensors are used for data acquisition, transmitting millimeter waves and measuring the millimeter wave signals reflected by the target. This typically involves setting sensor parameters, such as frequency and power, to ensure sufficient signal quality and weighting the reflected millimeter wave signal data using a window function. Window functions can be of various types, such as Hanning windows and Blackman windows. Weighting aims to smooth the signal in the time domain and reduce boundary effects. Signal segmentation is then performed on the weighted reflected millimeter wave signal data. This may involve applying a threshold or other segmentation algorithm to separate the signal into distinct segments. This helps separate the target's motion or features for better analysis. A discrete Fourier transform (DFT) or fast Fourier transform (FFT) is then performed on the segmented signal dataset. This converts the signal from the time domain to the frequency domain, yielding a complex spectrum. Practical implementation may involve performing a Fourier transform using appropriate algorithms and parameters, and performing frequency domain analysis on the frequency domain signal dataset, which may include finding the dominant frequency components, performing frequency domain filtering, or other frequency domain operations. This helps identify the target's frequency signature and spectrum information, and performs Doppler shift calculations on the frequency signal change data, which involves analyzing frequency changes in the spectrum to obtain target velocity information. This may include using correlation analysis or other techniques to calculate Doppler shift.

[0099] Preferably, step S2 includes the following steps:

[0100] Step S21: performing bandpass filtering on the dynamic hand target signal data to obtain prominent phase signal data;

[0101] Step S22: performing phase demodulation processing on the prominent phase signal data to obtain a complex amplitude signal;

[0102] Step S23: performing a differential operation on the complex amplitude signal to obtain a differential signal;

[0103] Step S24: Calculate the phase change according to the differential signal to obtain hand phase change information data.

[0104] The present invention uses bandpass filtering to filter out signal components within a specific frequency range and remove unnecessary frequency components. Here, it is used to highlight the phase information of the dynamic target signal, making subsequent processing more accurate. Phase demodulation converts the highlighted phase signal data into a complex amplitude signal, from which the phase information of the target is extracted. This helps to more clearly represent the target's motion, shape or other phase-related features. The differential operation is used to calculate the changes in the complex amplitude signal to capture the changes in the signal. This can enhance the perception of the target's motion or other dynamic features and facilitate further analysis. By calculating the phase changes of the differential signal, more detailed information about the dynamic target can be obtained. These hand phase change information may be related to the target's speed, acceleration or other motion characteristics, providing a more comprehensive description of the motion.

[0105] In an embodiment of the present invention, filters used in digital signal processing technology are used, such as Butterworth filters, FIR filters, and the like. Appropriate filter types and parameters are selected based on the frequency characteristics of the target signal to perform bandpass filtering operations, and phase demodulation techniques, such as Hilbert transform or demodulators, are used. These techniques can convert prominent phase signals into complex amplitude signals, where phase information is extracted and embedded in the complex amplitude signal. A differential operation is performed on the complex amplitude signal to calculate the difference between adjacent samples. This can be achieved by simply performing a differential operation on the signal, i.e., calculating the difference between adjacent samples and calculating the phase change based on the differential signal. This may include further processing of the differential signal, such as using a phase extraction algorithm or a differential operation, to obtain accurate hand phase change information.

[0106] Preferably, step S3 includes the following steps:

[0107] Step S31: using a phase gesture conversion algorithm to perform feature conversion on the hand phase change information data to obtain effective gesture feature data;

[0108] Step S32: performing gesture recognition processing on the valid gesture feature data using a machine learning algorithm to obtain operation gesture data;

[0109] Step S33: constructing a three-dimensional model based on the operation gesture data to obtain a reconstructed hand posture model;

[0110] Step S34: geometrically subdividing the reconstructed hand posture model to obtain an operation gesture model;

[0111] Step S35: cutting the operating hand model to obtain the hand-machine interaction area and part geometry.

[0112] The present invention uses a phase gesture conversion algorithm to convert hand phase change information into effective gesture feature data. Such feature data is more representative and can better capture the dynamic changes and special features of gestures. The effective gesture feature data is processed using a machine learning algorithm to achieve automatic recognition of gestures. This can be used to map gestures to specific control commands or interactions with the system, improving the interaction efficiency between users and the system. A three-dimensional model is built based on the operation gesture data, enabling the system to model hand gestures more realistically and accurately. This provides a higher level of information for subsequent processing. Geometric subdivision of the reconstructed hand gesture model can improve the resolution of gesture details, making the model more accurate and better capturing subtle changes in hand gestures. Model cutting of the operation hand model can extract the machine interaction area of ​​the hand and the geometric shape of each part. This helps the system understand the user's gesture movements more accurately and respond in a targeted manner.

[0113] As an example of the present invention, refer to Figure 2 As shown, in this example, step S3 includes:

[0114] Step S31: using a phase gesture conversion algorithm to perform feature conversion on the hand phase change information data to obtain effective gesture feature data;

[0115] In an embodiment of the present invention, the hand phase change information data is preprocessed to ensure data consistency and quality, including operations such as denoising, standardization, and normalization. An appropriate feature extraction method is selected, which may involve frequency domain features, time domain features, spatial domain features, etc. Common features include amplitude, frequency, phase difference, energy, etc. The extracted features are reduced in dimensionality to reduce computational complexity and avoid overfitting. Based on the problem requirements, an appropriate model is selected for feature conversion, including traditional signal processing methods and models based on deep learning, such as convolutional neural networks (CNNs) or recurrent neural networks (RNNs). A labeled gesture dataset is prepared and the model is trained. This may involve supervised learning or unsupervised learning. The trained model is evaluated to check its performance, and the trained model is applied to new hand phase change information data to obtain valid gesture feature data.

[0116] Step S32: performing gesture recognition processing on the valid gesture feature data using a machine learning algorithm to obtain operation gesture data;

[0117] In an embodiment of the present invention, appropriate features are selected or extracted based on the characteristics of the gesture data. This may include valid gesture feature data obtained from S31, or other features related to gestures. The features are standardized to ensure that the numerical ranges of different features are consistent to improve the stability of the model, and an appropriate machine learning model is selected. Common ones include support vector machines (SVM), decision trees, random forests, neural networks, etc. The selected model should match the nature of the gesture recognition task, and the selected machine learning model is trained using a training set. In this step, the model learns how to distinguish different gesture categories from the features, and the trained model is evaluated using a test set. Evaluation indicators may include accuracy, precision, recall, F1 score, etc. Based on the evaluation results, it may be necessary to adjust the model's hyperparameters or select a different model to improve performance. After the model training and evaluation are completed, the model can be used to predict new gesture data to obtain operation gesture data.

[0118] Step S33: constructing a three-dimensional model based on the operation gesture data to obtain a reconstructed hand posture model;

[0119] In embodiments of the present invention, an appropriate 3D modeling method is selected based on the nature of the gesture data. This may include point cloud-based, voxel-based, or deep learning-based methods to pre-process the gesture data to ensure data consistency and quality. This may include operations such as denoising, smoothing, and calibration to map the 2D or 3D coordinates in the gesture data to actual 3D space. This may require the use of techniques such as camera calibration to ensure accurate spatial mapping. A 3D model is then constructed from the mapped gesture data using the selected 3D modeling method. This may include operations such as point cloud fitting and surface reconstruction. If the specific positions of hand joints and hand posture information are required, joint detection and posture estimation methods may be used. This typically involves using a deep learning model, such as convolutional neural network-based hand keypoint detection, to optimize the constructed 3D model to improve its accuracy and realism. This may include removing outliers, adjusting model parameters, and visualizing the constructed 3D model for further inspection and analysis. This can be accomplished using 3D rendering techniques or other visualization tools to verify that the constructed 3D model conforms to the characteristics of the actual hand posture. This can be accomplished by comparing the model predictions with the actual gesture data.

[0120] Step S34: geometrically subdividing the reconstructed hand posture model to obtain an operation gesture model;

[0121] In embodiments of the present invention, an appropriate geometric subdivision method is selected based on the requirements and model characteristics. Geometric subdivision can be achieved using a variety of techniques, such as mesh subdivision and surface subdivision. The reconstructed hand gesture model is preprocessed to ensure that the model's topology and geometry are suitable for geometric subdivision. This may include triangulating the model, removing unnecessary portions, and ensuring surface consistency. The model is then subdivided using the selected geometric subdivision method. This can be achieved by increasing mesh density, refining surfaces, and other methods. If the geometric subdivision method allows, subdivision control parameters can be adjusted to control the degree of subdivision. This helps find the right balance between model detail and computational efficiency. After geometric subdivision, some model optimization may be required, such as removing redundant geometric details and correcting topological errors. This is done to verify that the subdivided model meets expectations and to make necessary adjustments and optimizations. This may require comparison with actual gesture data to ensure the model maintains its fidelity and accuracy after subdivision. The subdivided gesture model is then exported to the desired format for use in the application.

[0122] Step S35: cutting the operating hand model to obtain the hand-machine interaction area and part geometry;

[0123] In this embodiment of the present invention, key areas and parts for hand-machine interaction are defined. This may involve key areas such as the fingers, palm, and wrist. Depending on the application requirements, an appropriate model segmentation algorithm is selected. Model segmentation can be achieved using a variety of algorithms, such as geometry-based segmentation and volume-based segmentation. The algorithm should be selected based on the model's characteristics and application scenario. The hand model is segmented using the selected segmentation algorithm. To ensure that the segmentation operation accurately isolates the key areas and parts for hand-machine interaction, optimization of the segmented areas may be necessary after segmentation. This may include removing unnecessary parts, filling possible holes, and adjusting boundaries. Geometric information of the hand-machine interaction areas and parts is extracted from the segmented model. This information includes the shape, size, and relative position of each part. This is used to verify that the segmented results meet expectations and align with actual hand-machine interaction requirements. Necessary adjustments and optimizations are performed to ensure that the model's geometry meets design requirements. The segmented geometry of the hand-machine interaction areas and parts is then exported to the required format for use in the application.

[0124] Preferably, the phase gesture conversion algorithm in step S31 is specifically as follows:

[0125]

[0126] Among them, Y(t) represents the operation gesture data sequence, that is, the gesture features after the final conversion, X(t) represents the hand phase change information data sequence, that is, the input original gesture data, Δt represents the small time difference, which is the time interval used for the derivative function. is a derivative operator used to calculate the derivative of a function with respect to time. X(t+Δt) represents the hand phase change information data within a small time interval Δt after time t, where t represents time. It represents the angle feature of normalizing the hand phase change information data to [-1, 1]. π represents pi, which is used to convert the angle into radians.

[0127] The present invention constructs a phase gesture conversion algorithm. The Y(t) in the formula represents the final converted gesture feature, which is obtained by processing the input original gesture data X(t). X(t) is the hand phase change information data sequence, which represents the phase change of the gesture. The Δt in the formula represents the small time difference, which is used for the time interval when the derivative function is calculated. Through the derivative operator That is, by calculating the derivative of the function with respect to time, we can obtain the rate of change of gesture data over time. Represents the angular feature of hand phase change information data normalized to [-1, 1]. Where π represents pi, which is used to convert angles to radians. The overall idea of ​​the formula is to first normalize the hand phase change information, then take its square root, perform a derivative operation on it, and calculate the differential in the limit case to finally obtain the converted gesture feature Y(t). Overall, the formula in the present invention may be intended to capture certain instantaneous features of gesture data. By normalizing the phase information, performing trigonometric functions, logarithmic operations, and derivative operations, certain specific features of the gesture in a short period of time can be extracted. By analyzing the rate of change of the gesture data, more intuitive and interpretable gesture features can be extracted. This conversion method can capture the speed and trend of gesture changes, thereby facilitating accurate recognition and understanding of gestures. When using conventional phase gesture conversion algorithms in the field, the phase information in the hand model can be parsed to achieve understanding and recognition of gestures. By applying the phase gesture conversion algorithm of the present invention, the speed and trend of gesture changes can be more accurately captured, thereby achieving more accurate recognition and understanding of gestures. Phase gesture conversion can provide a natural, intuitive and contactless way to make human-computer interaction more flexible and improve user experience in various application scenarios.

[0128] Preferably, step S4 includes the following steps:

[0129] Step S41: extracting geometric features of the part geometry based on morphological analysis to obtain part structural feature data;

[0130] Step S42: performing mechanical structure analysis on the part structural characteristic data to obtain part mechanical performance data;

[0131] Step S43: performing specific feature detection on the part structure feature data to obtain various hand feature data, wherein the specific feature detection includes measuring finger length and palm width, observing muscle tissue and joint flexibility, and examining the hand bone structure;

[0132] Step S44: performing a grasping ability assessment based on the part mechanical performance data and the hand characteristic data to obtain the hand grasping ability data.

[0133] The present invention extracts the geometric features of parts through morphological analysis, which can help understand the morphological features of the hand structure, such as the length of the fingers, the width of the palm, etc., and provide basic data for subsequent analysis. By analyzing the mechanical structure of the parts, data on the mechanical properties of the hand can be obtained, such as muscle strength, range of joint motion, etc., which is very important for evaluating the function and adaptability of the hand. Through specific feature detection, including measuring the length of fingers and the width of the palm, observing the condition of muscle tissue and joint flexibility, and examining the skeletal structure of the hand, the physiological characteristics and health status of the hand can be fully understood, providing data support for personalized evaluation. Combining the mechanical performance data of the parts and the various feature data of the hand to evaluate the grasping ability can help evaluate the individual's hand function, identify potential health problems, and provide a reference basis for rehabilitation training and improvement of athletic ability. This comprehensive assessment of the hand takes into account different aspects of its characteristics, from geometric morphology to mechanical properties, and then to physiological characteristics and functional performance. It provides a comprehensive hand assessment. Through morphological analysis and mechanical structure analysis, the hand characteristic data is quantified, which helps to quantify the structure and performance of the hand and provides a basis for further comparison and analysis. Specific hand characteristic detection includes observations of finger length, palm width, muscle tissue, joint flexibility, and bone structure. This information is very important for understanding an individual's physiological state and potential physiological characteristics. The mechanical performance data of the parts and the physiological characteristic data of the hand are combined to assess grasping ability. This comprehensive assessment helps to understand the relationship between hand structure and function and provides a basis for improving grasping ability.

[0134] As an example of the present invention, refer to Figure 3 As shown, in this example, step S4 includes:

[0135] Step S41: extracting geometric features of the part geometry based on morphological analysis to obtain part structural feature data;

[0136] In an embodiment of the present invention, the acquired image is preprocessed, including steps such as denoising, contrast enhancement, and image smoothing, to ensure the accuracy and stability of subsequent analysis, and the hand or specific part is segmented from the image. This can be achieved through image segmentation techniques such as threshold processing and edge detection, and morphological analysis methods are used to extract geometric features from the segmented image. These may include contour features, area, perimeter, convex hull, skeleton, etc. Morphological operations such as dilation, erosion, opening, and closing can be used to adjust and optimize the image, quantify the extracted geometric features, and convert them into digital form for subsequent analysis and comparison. This may include calculating ratios, lengths, angles, etc., and storing the obtained part structural feature data for use in subsequent steps. This can be done using a database, text file, or other appropriate storage method to verify the extracted features and ensure their accuracy and reliability. If necessary, calibration is performed to suit specific research or application requirements.

[0137] Step S42: performing mechanical structure analysis on the part structural characteristic data to obtain part mechanical performance data;

[0138] In an embodiment of the present invention, part structural feature data is imported into a mechanical structural analysis tool or software. This may include information such as the size, shape, and mass distribution of geometric features. Finite Element Analysis (FEA) and other methods are then used to establish a mechanical model of the part. This involves decomposing the part into discrete finite elements and defining material properties, boundary conditions, and loading conditions. The model boundary conditions, including constraints and loading, are determined, and predetermined loading conditions are applied to the model to simulate the external forces to which the part is subjected during actual use or testing. This includes static loading, dynamic loading, or other specific mechanical loading. The model is solved using finite element analysis tools to obtain information such as deformation, stress, and strain of the part under load. These results provide quantitative data on the mechanical properties of the part. The model solution results are analyzed to obtain key mechanical performance parameters, such as maximum stress, deformation, and strain energy, which are used to evaluate the performance of the part in actual use. Based on the results of the mechanical analysis, design optimization is performed to improve the mechanical properties of the part, including adjustments to material selection, geometry, or structural design. The mechanical performance data of the part is then output and presented in a table, graph, or other format.

[0139] Step S43: performing specific feature detection on the part structure feature data to obtain various hand feature data, wherein the specific feature detection includes measuring finger length and palm width, observing muscle tissue and joint flexibility, and examining the hand bone structure;

[0140] In an embodiment of the present invention, the muscle tissue of the hand is observed, and any abnormal or prominent features are noted. The tension and texture of the muscles are assessed by visual inspection or light touch. The range of motion of the finger and wrist joints is tested, including flexion, extension, rotation and other movements. The joints are observed for abnormal stiffness and for mobility within the normal range. The skeletal structure of the hand is examined to observe whether the bones are normal and to check for fractures, deformities or other abnormalities. The gaps between the joints are analyzed to assess the health of the joints. The measurements, observations and imaging examination results are recorded to establish a detailed file of the hand structure. The records can be made in the form of charts, text descriptions or digital data. The various features of the hand structure are analyzed and compared with the standards within the normal range or with other reference data to identify any abnormal features that may affect the function or health of the hand.

[0141] Step S44: performing a grasping ability assessment based on the part mechanical performance data and the hand characteristic data to obtain the hand grasping ability data.

[0142] In an embodiment of the present invention, various hand feature data are integrated into the evaluation system, including information such as finger length, palm width, muscle tissue condition, joint flexibility and bone structure. According to the project requirements or research objectives, the evaluation indicators of grasping ability are determined, including factors such as grip strength, hand adaptability, and grip stability. Experiments or test plans are designed to ensure that the hand grasping ability can be comprehensively evaluated. Various grasping movements, such as fine grasping, coarse grasping, and side grasping, are used to simulate actual usage scenarios, and grasping ability tests are conducted according to the experimental plan. During the test, relevant mechanical performance data, such as grip curves and grip duration, are recorded, and a comprehensive analysis is performed in combination with the mechanical performance data and hand feature data. The relationship between specific features and grasping ability is explored, and possible influencing factors, such as hand anatomical structure and muscle strength, are considered. A grasping ability evaluation report is written to summarize the experimental results and conclusions.

[0143] Preferably, step S5 includes the following steps:

[0144] Step S51: Dynamically analyze the hand-machine interaction area to obtain action sequence data of the interaction area;

[0145] Step S52: performing action sequence comparison on the interactive area action sequence data to obtain comparative analysis data;

[0146] Step S53: extracting hand deformation information data from the comparative analysis data;

[0147] Step S54: Based on the preset hand force and deformation relationship, an inversion operation is performed on the hand deformation information data using a force and deformation inversion algorithm to obtain the hand applied force data.

[0148] The present invention dynamically analyzes the hand-machine interaction area, providing a comprehensive understanding of the hand's movements and postures during specific tasks or activities. This helps capture the diversity and complexity of hand movements. Motion sequence comparison allows for comparison of different motion sequences, generating comparative analysis data. This provides a foundation for quantitative analysis, identifying similar or different motion patterns and helping to understand differences in hand behavior in different contexts. Extracting hand deformation information from the comparative analysis data further refines the description of the hand's state. This includes information such as hand deformation and distortion, providing more detailed data for subsequent analysis. Based on a preset relationship between hand force and deformation, a force-deformation inversion algorithm is used to convert deformation information into force data. This preset relationship can be based on previous experimental data, biomechanical principles, or other domain expertise, making the inversion results more accurate and reliable. Using the force-deformation inversion algorithm, S54 accurately converts deformation information into hand-applied force data. This provides critical quantitative information for a deeper understanding of the mechanical properties of the hand during interaction. This method can be applied in many fields, including human-computer interaction design, rehabilitation engineering, and virtual reality technology. By gaining a deeper understanding of the dynamic behavior and forces exerted by the hand, we can better optimize the design and improve the user experience.

[0149] As an example of the present invention, refer to Figure 4 As shown, in this example, step S5 includes:

[0150] Step S51: Dynamically analyze the hand-machine interaction area to obtain action sequence data of the interaction area;

[0151] In the examples of the present invention, the collected raw data is preprocessed, including denoising, filtering, calibration and other operations, and the target tracking technology is used to track and record the movement trajectory of the hand in the interactive area to obtain the position and posture information of the hand. By analyzing the movement trajectory and posture information of the hand, the action sequence data of the hand in the interactive area is extracted, including the hand movement trajectory, the degree of bending of the fingers, the direction of the palm, etc., and the identified hand movements are combined into an action sequence. This involves arranging the actions in chronological order to form a time series for subsequent comparison and analysis, and extracting key action features from the action sequence. This includes the amplitude, speed, acceleration, duration, etc. of the action, and the processed dynamic analysis data is stored in an appropriate database or file for subsequent processing and analysis.

[0152] Step S52: performing action sequence comparison on the interactive area action sequence data to obtain comparative analysis data;

[0153] In an embodiment of the present invention, an appropriate similarity metric is selected to measure the similarity between two action sequences, and key action features are extracted therefrom, including the frequency, amplitude, speed, duration, etc. of the action. The action sequence data is standardized to ensure that the comparison between different data is meaningful. The action amplitude, timing, etc. are adjusted to eliminate the scale effect. A similarity calculation method, such as Euclidean distance, cosine similarity, etc., is used to calculate and compare the similarity between different action sequence data to obtain a similarity matrix. Based on the similarity matrix, a cluster analysis method is used to cluster the action sequence data, and similar action sequence data are divided into the same category for classification and identification. Based on the identification results, comparative analysis data is generated, the meaning of the comparative analysis data is interpreted, and the similarities or differences between different action sequences are understood. This involves the application of domain expertise to ensure the correct interpretation of the comparative analysis results.

[0154] Step S53: extracting hand deformation information data from the comparative analysis data;

[0155] In an embodiment of the present invention, hand deformation features that need to be paid attention to in comparative analysis are determined, including the hand's motion trajectory, the bending angle of the fingers, the deformation of the palm, etc., and the parts related to the hand deformation are screened out based on the previous comparative analysis data. This involves identifying the time points or sequences of hand movements in the comparative analysis data, parsing the data in the selected time points or sequences into specific hand deformation information, including coordinate positions, angle changes, shape changes, etc., ensuring that the data analysis is accurate and consistent with the selected features, and performing feature extraction on the parsed hand deformation data to simplify the data and capture key information. Feature extraction may include calculating the maximum deformation value, average angle, range of motion, etc. The extracted hand deformation information data is organized and stored, and the extracted hand deformation information is used for further data analysis, including statistical analysis, trend analysis, pattern recognition, etc., to draw deeper conclusions about hand deformation.

[0156] Step S54: Based on the preset hand force and deformation relationship, an inversion operation is performed on the hand deformation information data using a force and deformation inversion algorithm to obtain the hand applied force data.

[0157] In an embodiment of the present invention, a mathematical model of force and deformation is established based on a preset relationship between hand force and deformation. This involves using mechanical principles or machine learning methods to describe the relationship between hand deformation and applied force. An appropriate inversion algorithm is selected, which should be able to infer the corresponding applied force from hand deformation information. This includes numerical solution methods, optimization algorithms, or machine learning model training. The hand deformation information data is preprocessed to ensure data quality and consistency, including noise removal, missing value filling, and data standardization. The parameters required for the inversion algorithm are initialized, including initial parameters of the force and deformation model and the initial point of the optimization algorithm. The selected inversion algorithm is run, the hand deformation information is input into the model, and an attempt is made to find the optimal force and deformation relationship to invert the hand applied force data. Based on the actual situation, the inversion results are optimized and adjusted, including adjusting model parameters and inversion algorithm hyperparameters, to improve the accuracy of the inversion results. The obtained hand applied force data is verified and evaluated, including comparison with known force data, error calculation, or using methods such as cross-validation to interpret the inversion results and understand the relationship between hand deformation and force.

[0158] Preferably, the force and deformation inversion algorithm in step S54 is specifically as follows:

[0159]

[0160] Among them, F(x,a) is the force exerted by the hand at position x at time t, that is, the result obtained by inversion, and u(x,s) is the deformation information at position x, including displacement, deformation and other information. is a gradient operator used to calculate the gradient of the deformation field u(x,s), Ω is the overall deformation area of ​​the hand, and defines the distribution of the deformation information u(x,s) within the entire hand range, e -a is an exponential term, which represents time attenuation, indicating that the effect of force gradually decreases with the passage of time. cos(a) is a trigonometric function term, which represents the periodic force change at time a. T is the upper limit of time integration. is the partial derivative, is the partial derivative of the deformation function u with respect to the spatial coordinate s, is a weight or influence term based on the distance between points x and s, and ds is an infinitesimal element in the spatial domain Ω.

[0161] The present invention constructs a force and deformation inversion algorithm, in which the e -aThe time decay term, an exponential term that represents the time decay of the force field, is used to describe the force field's decay. As time a increases, the force gradually decreases, reflecting a time decay effect. cos(a) is the periodic force variation term. The trigonometric term cos(a) introduces a periodic variation, indicating that at time a, the force field undergoes a sinusoidal change. This can be used to simulate periodic forces applied by a hand. is the spatial weight term, which is a weight or influence term based on the distance between points x and s. It expresses the spatial trade-off when the deformation information is propagated to position x, where the closer the distance between the deformation information and position s is to x, the greater its influence. is the spatial gradient term, which represents the integral of the partial derivative of the deformation field u(x,s) with respect to the spatial coordinate s over the entire hand deformation area Ω, and is calculated through the gradient operator Calculation. This reflects the spatial distribution of deformation information and the influence of its changes on the force field. By mathematically modeling the deformation information of the hand, the force field changes in space and time are simulated. The interaction between the parameters causes the force field to gradually decrease over time. Taking into account the continuity and gradualness in time, a periodic sinusoidal waveform is introduced, so that the force field presents periodic fluctuations in time, which is suitable for simulating the periodic movements applied by the hand. Through the weight factor of the distance, the propagation of the deformation information in space is more reasonably considered, so that the points closer to the force field contribute more to the force field. The influence of the gradient of the deformation information on the force field is taken into account, and the changes in the deformation information in space are captured more comprehensively. When using the conventional force and deformation inversion algorithm in this field, the force and deformation inversion results can be obtained. By applying the force and deformation inversion algorithm provided by the present invention, the force and deformation inversion results can be calculated more accurately. The inversion algorithm can provide a stress distribution map inside the object, showing the magnitude and direction of the force on each part of the object, providing a detailed understanding of the stress and deformation inside the object, which helps to improve the design, optimize the performance and ensure the safety of the structure.

[0162] Preferably, step S6 includes the following steps:

[0163] Step S61: performing interaction force analysis based on the hand grasping ability data and the hand applied force data to obtain hand-machine interaction force data;

[0164] Step S62: Perform interaction mechanics analysis on the hand-machine interaction force data to obtain machine mastery data;

[0165] Step S63: performing interval positioning on the machine mastery data according to the preset machine mastery data interval to obtain the energy-saving mode interval;

[0166] Step S64: switching the energy-saving mode of the digital welding machine according to the energy-saving mode range.

[0167] By considering hand grip and applied force, the present invention enables the system to be personalized to the specific needs of users. By combining hand grip data and applied force data for interaction force analysis, a more comprehensive understanding of the forces applied by the hand during interaction with the machine can be achieved. This helps optimize machine design, improve the interactive interface, and enhance the effectiveness and safety of machine-user interaction. Interaction mechanics analysis of hand-machine interaction force data allows for in-depth study of the mechanical properties between the machine and hand, including force magnitude, direction, and point of application. This helps optimize the machine's mechanical structure and control algorithms, improving its stability and accuracy. Different users have different mechanical characteristics. Through personalized adaptation, digital welders can more effectively meet user work requirements. Interaction mechanics analysis helps understand the mechanical relationship between the hand and the machine. This understanding can be used to improve machine design, enhancing user interactivity and comfort. The generated machine mastery data provides quality information about machine usage. This can be used to evaluate machine performance and provide a strong basis for design improvements. By positioning within a preset range based on machine mastery data, the digital welder can be switched to an energy-saving mode. This can effectively reduce energy consumption and improve the energy efficiency of the equipment. By considering the mechanical characteristics of the user's hand, the system can better adapt to the user's habits and needs, thereby improving the user experience. This is especially important for operators who use digital welders for a long time. By running the digital welder in energy-saving mode, energy use can be effectively reduced, operating costs can be reduced, and a positive impact on the environment can be achieved. By comprehensively considering the user's physiological characteristics and the working status of the machine, personalized and efficient digital welder operation is achieved, and energy consumption is reduced while improving user satisfaction.

[0168] In an embodiment of the present invention, hand gripping ability data and hand applied force data are obtained, and the collected hand gripping ability and applied force data are integrated to establish a comprehensive model of hand-machine interaction force. The established model is used to perform interaction force analysis, identify the interaction force between the user's hand and the machine when using the digital welding machine, and generate hand-machine interaction force data, which describes the force applied by the user to the machine when using the digital welding machine. The hand-machine interaction force data is input into an interactive mechanics analysis system, and the hand-machine interaction force data is analyzed in detail using appropriate mathematical and physical tools, such as Newtonian mechanics or other mechanics principles. Based on the analysis results, data on machine mastery is obtained, wherein the machine mastery data includes information on the stability of the machine, the user's mastery of the machine, and other aspects. The machine mastery data is generated and used to evaluate the performance and user experience of the digital welding machine during use, and an interval of the machine mastery data is preset in the system. The preset intervals of machine mastery data are set based on previous user research, ergonomic principles or other relevant standards. The machine mastery data is mapped to the preset intervals to determine the current mastery level of the digital welder. Different energy-saving modes are defined in the digital welder, which may involve power adjustment, standby mode or other energy-saving strategies. Based on the results of interval positioning, it is determined whether the current machine mastery level is within the preset energy-saving mode interval. If the mastery level is within the energy-saving mode interval, it triggers switching to the corresponding energy-saving mode to reduce the energy consumption of the digital welder, provide user feedback, and notify the user that the digital welder has switched to energy-saving mode to ensure that the user understands the status changes of the system. During the use of the digital welder, the system can monitor the interaction force and machine mastery in real time, and continuously optimize the energy-saving mode switching strategy to adapt to the changing needs of the user.

[0169] In this specification, a digital welding machine energy-saving mode switching system is provided, which is used to execute the above-mentioned digital welding machine energy-saving mode switching method, including:

[0170] The hand acquisition module is used to obtain the millimeter wave signal data reflected by the digital welder; perform Doppler frequency shift processing on the millimeter wave signal data reflected by the digital welder to obtain dynamic hand target signal data;

[0171] The data processing module is used to perform phase demodulation processing on the dynamic hand target signal data to obtain a complex amplitude signal; and perform phase estimation processing on the complex amplitude signal to obtain hand phase change information data;

[0172] The model reconstruction module is used to perform feature conversion processing on the hand phase change information data to obtain the operation gesture data; use the operation gesture data to reconstruct the operation hand posture to obtain the operation hand model; and perform model cutting on the operation hand model to obtain the hand-machine interaction area and part geometry;

[0173] The ability assessment module is used to perform morphological analysis on the geometric shape of the part to obtain the structural feature data of the part; and to perform grasping ability assessment based on the structural feature data of the part to obtain the hand grasping ability data;

[0174] The deformation inversion module is used to perform dynamic analysis on the hand-machine interaction area to obtain hand deformation information data; based on the preset hand force and deformation relationship, the hand deformation information data is inverted to obtain the hand applied force data;

[0175] The mode switching module is used to perform interactive force analysis on the hand grasping ability data and the hand applied force data to obtain the hand-machine interactive force data; and switch the energy-saving mode of the digital welding machine according to the hand-machine interactive force data.

[0176] The present invention utilizes reflected millimeter wave signals for data acquisition, eliminating the need for direct hand contact and avoiding the limitations of contact measurement with traditional sensors. Through phase demodulation and estimation, the dynamic target signal is converted into a complex amplitude signal. Hand phase change information is extracted and feature conversion is performed, enabling recognition of hand gestures and reconstruction of hand posture. Hand models are then segmented and subjected to morphological analysis of part geometry to obtain part structural feature data, further enabling assessment of the hand's grasping ability. This helps understand the hand's function and adaptability to different operations. Dynamic analysis of the hand-machine interaction area and deformation information data allows estimation of the force applied by the hand. Based on a preset hand force-deformation relationship, an inversion operation is performed to obtain hand-applied force data, providing a more comprehensive understanding of the force interaction between the hand and the object. Combined with the hand's grasping ability data and hand-applied force data, interaction force analysis is performed to obtain more comprehensive hand-machine interaction force data. This helps assess the hand's needs and adaptability to digital welding machine operations. Using this hand-machine interaction force data, the digital welding machine's energy-saving mode can be switched according to pre-defined rules or algorithms. In this way, the energy consumption of the digital welding machine can be adjusted according to the actual needs and capabilities of the hand, thereby improving energy utilization efficiency. The present invention can obtain hand-related information in a non-contact manner, and combine hand posture, grasping ability and machine interaction force data to provide a more comprehensive basis for switching the energy-saving mode of the digital welding machine, thereby achieving more intelligent and efficient energy consumption management.

[0177] The present invention is therefore intended to be illustrative and non-restrictive in all respects, with the scope of the invention being defined by the appended claims rather than the foregoing description, and all changes that come within the meaning and range of equivalents of the application documents are intended to be embraced therein.

[0178] The foregoing description is intended only to provide specific embodiments of the present invention, which will enable those skilled in the art to understand and implement the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention is not intended to be limited to the embodiments shown herein, but is to be construed in the widest possible manner consistent with the principles and novel features disclosed herein.

Claims

1. A method for switching energy-saving modes of a digital welding machine, characterized in that: The following steps are involved: Step S1: obtaining millimeter wave signal data reflected by a digital welder; performing Doppler frequency shift processing on the millimeter wave signal data reflected by the digital welder to obtain dynamic hand target signal data; Step S2: performing phase demodulation processing on the dynamic hand target signal data to obtain a complex amplitude signal; performing phase estimation processing on the complex amplitude signal to obtain hand phase change information data; Step S3: performing feature conversion processing on the hand phase change information data to obtain operation gesture data; reconstructing the operation hand posture using the operation gesture data to obtain an operation hand model; and performing model segmentation on the operation hand model to obtain the hand-machine interaction area and part geometry; Step S4: performing morphological analysis on the geometrical morphology of the part to obtain part structural characteristic data; performing grasping ability evaluation based on the part structural characteristic data to obtain hand grasping ability data; Step S5: Dynamically analyzing the hand-machine interaction area to obtain hand deformation information data; performing an inversion operation on the hand deformation information data based on a preset hand force and deformation relationship to obtain hand applied force data; Step S6: performing interaction force analysis on the hand gripping ability data and the hand applied force data to obtain hand-machine interaction force data; and switching the energy-saving mode of the digital welding machine according to the hand-machine interaction force data.

2. The method for switching the energy-saving mode of a digital welding machine according to claim 1, characterized in that: Step S1 includes the following steps: Step S11: using a millimeter wave sensor to collect data from the digital welder to obtain millimeter wave signal data reflected by the digital welder; Step S12: using a window function to perform weighted processing on the digital welding machine reflected millimeter wave signal data to obtain weighted reflected millimeter wave signal data; Step S13: performing signal segmentation on the weighted reflected millimeter wave signal data to obtain a segmented signal data set; Step S14: performing Fourier transform on the segmented signal data set to obtain a frequency domain signal data set; Step S15: performing frequency domain analysis on the frequency domain signal data set to obtain frequency signal change data; Step S16: performing Doppler frequency shift calculation on the frequency signal change data to obtain dynamic hand target signal data.

3. The method for switching energy-saving mode of a digital welding machine according to claim 1, characterized in that: Step S2 includes the following steps: Step S21: performing bandpass filtering on the dynamic hand target signal data to obtain prominent phase signal data; Step S22: performing phase demodulation processing on the prominent phase signal data to obtain a complex amplitude signal; Step S23: performing a differential operation on the complex amplitude signal to obtain a differential signal; Step S24: Calculate the phase change according to the differential signal to obtain hand phase change information data.

4. The method for switching the energy-saving mode of a digital welding machine according to claim 1, characterized in that: Step S3 includes the following steps: Step S31: using a phase gesture conversion algorithm to perform feature conversion on the hand phase change information data to obtain effective gesture feature data; Step S32: performing gesture recognition processing on the valid gesture feature data using a machine learning algorithm to obtain operation gesture data; Step S33: constructing a three-dimensional model based on the operation gesture data to obtain a reconstructed hand posture model; Step S34: geometrically subdividing the reconstructed hand posture model to obtain an operation gesture model; Step S35: cutting the operating hand model to obtain the hand-machine interaction area and part geometry.

5. The method for switching the energy-saving mode of a digital welding machine according to claim 4, characterized in that: The phase gesture conversion algorithm in step S31 is as follows: Among them, Y(t) represents the operation gesture data sequence, that is, the gesture features after the final conversion, X(t) represents the hand phase change information data sequence, that is, the input original gesture data, Δt represents the small time difference, which is the time interval used for the derivative function. is a derivative operator used to calculate the derivative of a function with respect to time. X(t+Δt) represents the hand phase change information data within a small time interval Δt after time t, where t represents time. It represents the angle feature of normalizing the hand phase change information data to [-1, 1]. π represents pi, which is used to convert the angle into radians.

6. The method for switching the energy-saving mode of a digital welding machine according to claim 1, characterized in that: Step S4 includes the following steps: Step S41: extracting geometric features of the part geometry based on morphological analysis to obtain part structural feature data; Step S42: performing mechanical structure analysis on the part structural characteristic data to obtain part mechanical performance data; Step S43: performing specific feature detection on the part structure feature data to obtain various hand feature data, wherein the specific feature detection includes measuring finger length and palm width, observing muscle tissue and joint flexibility, and examining the hand bone structure; Step S44: performing a grasping ability assessment based on the part mechanical performance data and the hand characteristic data to obtain the hand grasping ability data.

7. The method for switching energy-saving modes of a digital welding machine according to claim 1, characterized in that: Step S5 includes the following steps: Step S51: Dynamically analyze the hand-machine interaction area to obtain action sequence data of the interaction area; Step S52: performing action sequence comparison on the interactive area action sequence data to obtain comparative analysis data; Step S53: extracting hand deformation information data from the comparative analysis data; Step S54: Based on the preset hand force and deformation relationship, an inversion operation is performed on the hand deformation information data using a force and deformation inversion algorithm to obtain the hand applied force data.

8. The method for switching the energy-saving mode of a digital welding machine according to claim 7, characterized in that: The force and deformation inversion algorithm in step S54 is as follows: Among them, F(x,a) is the force exerted by the hand at position x at time t, that is, the result obtained by inversion, and u(x,s) is the deformation information at position x, including displacement, deformation and other information. is a gradient operator used to calculate the gradient of the deformation field u(x,s), Ω is the overall deformation area of ​​the hand, and defines the distribution of the deformation information u(x,s) within the entire hand range, e -a is an exponential term, which represents time attenuation, indicating that the effect of force gradually decreases with the passage of time. cos(a) is a trigonometric function term, which represents the periodic force change at time a. T is the upper limit of time integration. is the partial derivative, is the partial derivative of the deformation function u with respect to the spatial coordinate s, is a weight or influence term based on the distance between points x and s, and ds is an infinitesimal element in the spatial domain Ω.

9. The method for switching energy-saving modes of a digital welding machine according to claim 1, characterized in that: Step S6 includes the following steps: Step S61: performing interaction force analysis based on the hand grasping ability data and the hand applied force data to obtain hand-machine interaction force data; Step S62: Perform interaction mechanics analysis on the hand-machine interaction force data to obtain machine mastery data; Step S63: performing interval positioning on the machine mastery data according to the preset machine mastery data interval to obtain the energy-saving mode interval; Step S64: switching the energy-saving mode of the digital welding machine according to the energy-saving mode range.

10. A method and system for switching energy-saving modes of a digital welding machine, characterized in that: For executing the digital welding machine energy-saving mode switching method according to claim 1, the digital welding machine energy-saving mode switching system comprises: The hand acquisition module is used to obtain the millimeter wave signal data reflected by the digital welder; perform Doppler frequency shift processing on the millimeter wave signal data reflected by the digital welder to obtain dynamic hand target signal data; The data processing module is used to perform phase demodulation processing on the dynamic hand target signal data to obtain a complex amplitude signal; and perform phase estimation processing on the complex amplitude signal to obtain hand phase change information data; The model reconstruction module is used to perform feature conversion processing on the hand phase change information data to obtain the operation gesture data; use the operation gesture data to reconstruct the operation hand posture to obtain the operation hand model; and perform model cutting on the operation hand model to obtain the hand-machine interaction area and part geometry; The ability assessment module is used to perform morphological analysis on the geometric shape of the part to obtain the structural feature data of the part; and to perform grasping ability assessment based on the structural feature data of the part to obtain the hand grasping ability data; The deformation inversion module is used to perform dynamic analysis on the hand-machine interaction area to obtain hand deformation information data; based on the preset hand force and deformation relationship, the hand deformation information data is inverted to obtain the hand applied force data; The mode switching module is used to perform interactive force analysis on the hand grasping ability data and the hand applied force data to obtain the hand-machine interactive force data; and switch the energy-saving mode of the digital welding machine according to the hand-machine interactive force data.

Images