Robotic display terminal and method thereof

Abstract

This application provides a robot display terminal and its method, relating to the field of human-computer interaction. It comprises a display terminal and an adapter. The display terminal includes a screen, function buttons, a safety switch, LED indicators, and an encoder knob. The adapter includes independently operating VGA and RS232 interfaces for receiving video signals from the controller and for data communication, respectively. The adapter also has a built-in dynamic touch calibration module, which can correct the touch coordinate offset of the display screen based on ambient temperature and humidity. This effectively ensures the accuracy and reliability of touch operation of the robot display terminal in different environments, thereby improving user experience and guaranteeing stable device performance.

Description

Technical Field

[0001] This application relates to the field of human-computer interaction, and more specifically, to a robot display terminal and its method. Background Technology

[0002] As a core component of human-machine interaction in industrial robot systems, robot display terminals are widely used in intelligent manufacturing, logistics and warehousing, and other scenarios. Their touch accuracy directly impacts operational efficiency and equipment reliability. However, in industrial settings, robots often face complex environmental disturbances (such as sudden temperature changes and humidity fluctuations), causing the display screen to deform due to thermal expansion and contraction or moisture absorption, leading to touch coordinate shifts. Traditional touch calibration technologies often rely on static calibration or single-point compensation, which cannot dynamically respond to temporal changes in environmental parameters. Over long-term operation, significant errors accumulate, especially under high temperature and humidity or rapid temperature change conditions, resulting in prominent issues such as false touch triggering and positioning drift.

[0003] Therefore, an optimized robot display terminal is desired. Summary of the Invention

[0004] To address the aforementioned technical problems, this application is proposed. Embodiments of this application provide a robot display terminal and its method.

[0005] According to one aspect of this application, a robot display terminal is provided, comprising: a display terminal and an adapter; the display terminal includes a display screen, multiple function buttons, a safety switch, LED indicators, and an encoder knob; the adapter includes an interface module. The interface module includes a VGA interface and an RS232 interface. The VGA interface is used to receive video signals from the controller; the RS232 interface is used to realize data communication with the controller. The VGA interface and the RS232 interface are independent of each other.

[0006] According to another aspect of this application, an adaptive calibration method for a robot display terminal is provided, comprising: A time sequence of environmental parameters collected by a temperature and humidity sensor is obtained, the environmental parameters including ambient temperature and ambient humidity; Acquire the original touch signal and determine the original touch coordinates based on the mapping relationship between the original touch signal and the theoretical coordinates; The time sequence of the environmental parameters is encoded and decoded to obtain the lateral deformation coordinate compensation amount and the longitudinal deformation coordinate compensation amount; Based on the lateral deformation coordinate compensation amount and the longitudinal deformation coordinate compensation amount, the original touch coordinates are corrected to obtain the corrected touch coordinates.

[0007] Compared with existing technologies, the robot display terminal and method provided in this application consist of a display terminal and an adapter. The display terminal includes a display screen, function buttons, a safety switch, LED indicators, and an encoder knob. The adapter includes independently operating VGA and RS232 interfaces, used for receiving video signals from the controller and for data communication, respectively. The adapter also has a built-in dynamic touch calibration module, which can correct the touch coordinate offset of the display screen based on ambient temperature and humidity. This effectively ensures the accuracy and reliability of touch operation of the robot display terminal in different environments, thereby improving user experience and guaranteeing stable device performance. Attached Figure Description

[0008] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. In the drawings: Figure 1 This is a block diagram of a dynamic touch calibration module in a robot display terminal according to an embodiment of this application.

[0009] Figure 2 This is a block diagram of the deformation coordinate compensation calculation unit in the robot display terminal according to an embodiment of this application.

[0010] Figure 3 This is a block diagram of a first-level sub-unit for deep co-coding of environmental parameters in a robot display terminal according to an embodiment of this application.

[0011] Figure 4 This is a flowchart of an adaptive calibration method for a robot display terminal according to an embodiment of this application.

[0012] Figure 5 This is a front structural diagram of a robot display terminal according to other embodiments of this application.

[0013] Figure 6 This is a schematic diagram of the rear structure of a robot display terminal according to other embodiments of this application. Detailed Implementation

[0014] Hereinafter, exemplary embodiments according to this application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this application, and not all embodiments of this application. It should be understood that this application is not limited to the exemplary embodiments described herein.

[0015] As a key component of human-machine interaction in industrial robot systems, robot display terminals are widely used in smart manufacturing, logistics and warehousing, and other application scenarios. Their touch accuracy directly impacts operational efficiency and equipment reliability. However, in industrial environments, robots often need to cope with complex external disturbances (such as drastic temperature changes and humidity fluctuations). These factors can cause touch coordinate shifts in the display screen due to thermal expansion and contraction or moisture absorption deformation of the materials. Traditional touch calibration technologies typically rely on static calibration or single-point compensation, which struggles to dynamically adapt to changes in environmental parameters. With long-term equipment operation, errors gradually accumulate, especially under extreme conditions such as high temperature and humidity or rapid temperature changes, where touch misoperation and positioning drift become particularly pronounced, severely affecting equipment stability and user experience.

[0016] To address the aforementioned technical problems, this application proposes a robot display terminal, comprising: a display terminal and an adapter; the display terminal includes a display screen, multiple function buttons, a safety switch, LED indicators, and an encoder knob; the adapter includes an interface module; the interface module includes a VGA interface and an RS232 interface, the VGA interface being used to receive video signals from a controller; the RS232 interface being used to achieve data communication with the controller; wherein the VGA interface and the RS232 interface are independent of each other.

[0017] Correspondingly, in the aforementioned robot display terminal, environmental interference (temperature and humidity changes) may cause physical deformation of the touchscreen, requiring dynamic calibration algorithms to compensate for errors. However, existing technologies typically rely solely on single-dimensional compensation based on temperature or humidity, neglecting the nonlinear impact of their combined effect on material deformation. Furthermore, existing methods, based on fixed thresholds or offline calibrated compensation coefficients, struggle to adapt to continuous deformation characteristics in dynamic environments and fail to consider the temporal cumulative effects of environmental parameters (such as gradual humidity changes superimposed with temperature jumps), leading to compensation lag or overshoot. Therefore, in the aforementioned robot display terminal, the adapter also includes a dynamic touch calibration module. This module is used to correct the touch coordinate offset of the display screen based on ambient temperature and humidity. In other words, during dynamic touch calibration, it can predict the physical deformation trend of the screen in real time and dynamically compensate for touch coordinate errors through the extraction of temporal features of temperature and humidity and collaborative modeling of deformation effects, overcoming the bottleneck of poor adaptability of traditional static calibration to dynamic environments.

[0018] Specifically, the technical concept of this application involves continuously collecting time-series data of environmental parameters using temperature and humidity sensors, and combining this data with the original coordinates of the touch signal to construct a dynamic mapping relationship between environment, deformation, and touch error. First, the temperature and humidity time-series data are separated into independent time-series signals according to the parameter dimension. A one-dimensional convolutional network is used to extract the local temporal correlation features of temperature and humidity, capturing dynamic fluctuation patterns such as temperature jumps and gradual humidity changes. Then, deep collaborative analysis of environmental parameters is used to fuse the temporal features of temperature and humidity, analyzing their nonlinear coupling effect on screen deformation—for example, when high temperature and high humidity are superimposed, the material expansion coefficient may exhibit an exponential increase, while a single parameter change results in a linear response. Further, a sequence decoder is used to generate lateral and longitudinal deformation compensation amounts. This sequence decoder can learn the cumulative effect of environmental parameters (such as material creep caused by prolonged high temperatures) and dynamically adjust the compensation strategy, avoiding the lag of traditional linear interpolation methods. Finally, the system superimposes the compensation amounts onto the original touch coordinates, forming a closed-loop feedback correction to achieve real-time elimination of touch errors under environmental disturbances.

[0019] Specifically, Figure 1 This is a block diagram of a dynamic touch calibration module in a robot display terminal according to an embodiment of this application. Figure 1 As shown, the dynamic touch calibration module 100 includes: an environmental parameter data acquisition unit 110, used to acquire a time sequence of environmental parameters collected by a temperature and humidity sensor, the environmental parameters including ambient temperature and ambient humidity; an original touch coordinate determination unit 120, used to acquire an original touch signal and determine the original touch coordinates based on the mapping relationship between the original touch signal and theoretical coordinates; a deformation coordinate compensation calculation unit 130, used to perform sequence encoding and decoding on the time sequence of environmental parameters to obtain lateral deformation coordinate compensation and longitudinal deformation coordinate compensation; and a touch coordinate correction unit 140, used to correct the original touch coordinates based on the lateral deformation coordinate compensation and the longitudinal deformation coordinate compensation to obtain corrected touch coordinates.

[0020] In this embodiment, the environmental parameter data acquisition unit 110 is used to acquire a time sequence of environmental parameters collected by a temperature and humidity sensor, including ambient temperature and ambient humidity. It should be understood that the time sequence of environmental parameters collected by the sensor specifically includes actual measured values ​​of temperature and humidity, temperature and humidity change trends, and temperature and humidity change rates. Correspondingly, changes in ambient temperature cause thermal expansion and contraction of the display screen material. Generally, as temperature increases, the material expands, potentially causing a positive shift in the horizontal and vertical coordinates of the touch screen; as temperature decreases, the material contracts, potentially causing a negative shift in the coordinates. The faster and larger the rate and magnitude of temperature change, the greater the potential deformation of the material. Changes in ambient humidity cause the display screen material to absorb or release moisture, resulting in deformation. Increased humidity causes the material to absorb moisture and expand, potentially causing a positive shift in the horizontal and vertical coordinates; decreased humidity causes the material to release moisture and contract, potentially causing a negative shift in the coordinates. The trend and rate of humidity change also affect the deformation of the material. In summary, the time sequence of environmental parameters collected by the temperature and humidity sensor reflects the temporal changes of the environmental parameters. The cumulative effects of environmental parameters over time can have a continuous impact on display materials. For example, a gradual change in humidity combined with a sudden temperature change can cause the material to deform to varying degrees at different points in time. By conducting in-depth analysis of this time-series data, we can capture the dynamic change patterns of environmental parameters, thus laying the foundation for touch coordinate compensation and correction.

[0021] In this embodiment, the original touch coordinate determination unit 120 is used to acquire the original touch signal and determine the original touch coordinates based on the mapping relationship between the original touch signal and theoretical coordinates. It should be understood that the original touch signal is the direct signal generated when the user performs a touch operation on the display screen. Determining the original touch coordinates through this touch signal can accurately reflect the user's true operational intention. Specifically, before calibration, although there may be coordinate offsets due to environmental factors, the original touch coordinates represent the position the user actually wants to touch. Calibration based on these coordinates ensures that the calibrated coordinates better match the user's operational needs, thereby improving the accuracy and efficiency of human-computer interaction.

[0022] The following is a detailed explanation of a specific implementation process for "acquiring the original touch signal and determining the original touch coordinates based on the mapping relationship between the original touch signal and theoretical coordinates": Firstly, at the hardware level, the display screen of the robot's display terminal is equipped with touch sensing devices, commonly capacitive touch sensor arrays and resistive touch films. Capacitive touch sensors detect touch actions by utilizing the characteristic that the capacitance of the screen surface changes when a human touches it. When a user's finger approaches or touches the screen, it causes a slight change in the screen's capacitance. The sensor can sensitively capture these changes and convert them into electrical signals related to the touch location. Resistive touch films, on the other hand, acquire touch information by measuring the change in resistance during a touch. When a touch occurs, the resistance value at the touch point changes, thereby generating a corresponding electrical signal. These electrical signals initially record the touch location information.

[0023] Touch sensors acquire analog signals, which need to be converted into digital signals for subsequent processing and analysis. Analog-to-digital converters (ADCs) perform this crucial task, converting the analog touch signals into discrete digital data that precisely represents the various parameters at the time of the touch. The converted digital signals are then stably transmitted to the terminal's microprocessor via transmission lines, providing the data foundation for subsequent coordinate calculations.

[0024] At the software level, the system predefines a theoretical coordinate system. This coordinate system is closely related to the physical size and resolution of the display screen. Taking a common 1920×1080 resolution display screen as an example, its upper left corner is set as the origin (0, 0), and the coordinates of the lower right corner are (1920, 1080). This theoretical coordinate system provides a unified reference standard for determining the original touch coordinates, giving subsequent coordinate calculations a clear direction.

[0025] To establish the connection between the raw touch signal and theoretical coordinates, determining their mapping relationship is crucial. During the production of display terminals, a series of rigorous calibration procedures are employed to obtain this critical mapping relationship. Specifically, standard touch tools are used to perform touch operations at multiple specific locations on the display screen, such as the four corners and the center. With each touch, the system simultaneously records the signal value generated by the touch sensor and the actual coordinates of the touch point in the theoretical coordinate system. Through in-depth analysis and complex calculations of a large amount of such data, a conversion formula or lookup table between the touch signal value and the theoretical coordinates is ultimately obtained. This conversion formula or lookup table is the core basis for realizing the conversion from raw touch signal to raw touch coordinates.

[0026] When the microprocessor receives touch signal data from the touch sensor, it performs calculations and conversions based on a pre-defined mapping relationship. If a conversion formula is used, the microprocessor substitutes the touch signal value into a carefully derived formula, precisely calculating the corresponding x and y coordinates to determine the touch point's position in the theoretical coordinate system. If a lookup table is used, the microprocessor quickly searches for the corresponding theoretical coordinate value in a pre-built lookup table based on the received touch signal value. Through this calculation or lookup process, the original touch coordinates are finally obtained based on the original touch signal. These original touch coordinates reflect the actual position information of the user's touch on the display screen.

[0027] In this embodiment, the deformation coordinate compensation calculation unit 130 is used to perform sequence encoding and decoding on the time queue of the environmental parameters to obtain the lateral deformation coordinate compensation and the longitudinal deformation coordinate compensation. Specifically, Figure 2 This is a block diagram of the deformation coordinate compensation calculation unit in a robot display terminal according to an embodiment of this application. Figure 2 As shown, the deformation coordinate compensation calculation unit 130 includes: an environmental parameter data segmentation first-level subunit 131, used to segment the environmental parameter time queue according to the environmental parameter sample dimension to obtain the environmental temperature time queue and the environmental humidity time queue; an environmental parameter temporal encoding first-level subunit 132, used to perform one-dimensional convolution-based sequence encoding on the environmental temperature time queue and the environmental humidity time queue to obtain the environmental temperature temporal correlation feature encoding vector and the environmental humidity temporal correlation feature encoding vector; an environmental parameter deep co-encoding first-level subunit 133, used to input the environmental temperature temporal correlation feature encoding vector and the environmental humidity temporal correlation feature encoding vector into the environmental parameter deep co-analysis network to obtain the environmental parameter-deformation influence deep co-encoding vector; and a deformation coordinate compensation generation first-level subunit 134, used to input the environmental parameter-deformation influence deep co-encoding vector into the RNN-based sequence decoder to obtain the lateral deformation coordinate compensation amount and the longitudinal deformation coordinate compensation amount.

[0028] In this embodiment, the environmental parameter data segmentation first-level subunit 131 is used to segment the time queue of the environmental parameters according to the environmental parameter sample dimension to obtain the time queue of environmental temperature and the time queue of environmental humidity. Accordingly, considering the differentiated effects of temperature and humidity on screen deformation and their temporal correlation characteristics in the dynamic touch calibration process, temperature changes often cause instantaneous thermal expansion and contraction of screen materials (such as rapid deformation of the metal frame), while humidity penetration may cause slow moisture absorption and expansion of the polymer layer. Although both affect touch accuracy, their physical response speed and nonlinearity differ significantly. Directly inputting the mixed temperature and humidity time-series data into the model would lead to feature confusion, making it difficult to distinguish the dominant influence range of key parameters (e.g., the instantaneous effect of high temperature abrupt changes and the long-term effect of humidity accumulation). Therefore, in the technical solution of this application, the time queue of the environmental parameters is segmented according to the environmental parameter sample dimension to obtain the time queue of environmental temperature and the time queue of environmental humidity. By segmenting data according to environmental parameter samples, the system can establish independent time-series signal streams for temperature and humidity, preserving their respective temporal evolution patterns (such as the steep waveform of temperature steps and the smooth trend of gradual humidity changes). In other words, the separated temperature and humidity time-series data avoid mutual interference between parameters, enabling the model to specifically learn temperature-dominated rapid deformation characteristics (such as metal frame expansion) and humidity-dominated slow deformation characteristics (such as polymer layer creep).

[0029] In this embodiment, the environmental parameter temporal encoding first-level subunit 132 is used to perform one-dimensional convolution-based sequence encoding on the time queues of environmental temperature and environmental humidity to obtain the environmental temperature temporal correlation feature encoding vector and the environmental humidity temporal correlation feature encoding vector. It should be understood that using a one-dimensional convolutional network for sequence encoding aims to extract local temporal correlation features of environmental temperature and environmental humidity parameters. Specifically, as the one-dimensional convolutional kernel slides along the time axis, it can efficiently capture short-term fluctuation patterns (such as the pulse characteristics of a sudden temperature rise) and long-term gradual trends (such as the slope characteristics of a continuous increase in humidity). Simultaneously, through multi-layer convolution stacking, it gradually abstracts higher-order temporal dependencies (such as the superposition effect of periodic temperature fluctuations and monotonically increasing humidity).

[0030] In this embodiment, the first-level subunit 133 of the deep co-coding of environmental parameters is used to perform deep co-analysis of environmental parameters on the environmental temperature time-series correlation feature encoding vector and the environmental humidity time-series correlation feature encoding vector to obtain the environmental parameter-deformation influence deep co-coding vector. Specifically, Figure 3 This is a block diagram of a first-level sub-unit for deep co-coding of environmental parameters in a robot display terminal according to an embodiment of this application. Figure 3As shown, the first-level subunit 133 of the deep co-coding of environmental parameters includes: a time-series feature autocorrelation matrix submodule 1331, used to perform time-series feature amplification and autocorrelation decomposition based on autocorrelation matrix time-series dependency amplification and autocorrelation decomposition on the time-series features of the environmental temperature time-series correlation feature encoding vector and the environmental humidity time-series correlation feature encoding vector; a dual-granularity co-coding feature calculation and perturbation correction submodule 1332, used to perform parallel calculation of the synergistic effect of feature granularity and feature value granularity of the core features of temperature and humidity and perform perturbation correction; and a corrected co-coding feature fusion submodule 1333, used to fuse the co-coding vectors after dual-granularity modeling and perturbation correction to obtain the deep co-coding vector of environmental parameters-deformation influence.

[0031] It is understandable that the coupling effect of temperature and humidity on screen deformation in industrial scenarios exhibits high-order nonlinearity and dynamic time-varying characteristics. For example, when a sudden increase in temperature is superimposed with a gradual change in humidity, the screen material may undergo non-additive deformation due to the synergistic effect of thermal expansion and hygroscopic expansion. Traditional linear superposition models cannot accurately characterize such complex interactions. Furthermore, the time-series features of temperature and humidity often contain a large amount of redundant noise (such as instantaneous sensor jitter and short-term fluctuations in environmental parameters). Directly performing shallow feature fusion will introduce interference signals, leading to deformation prediction errors. Therefore, in the technical solution of this application, the environmental temperature time-series associated feature encoding vector and the environmental humidity time-series associated feature encoding vector are further subjected to deep co-analysis of environmental parameters to obtain the deep co-coding vector of environmental parameters-deformation influence.

[0032] In order to realize the transformation from the input temperature and humidity time series features to the output co-influence encoding vector, in this embodiment of the application, a deep co-analysis network based on dual-granularity interaction and cross-granularity manifold correction is adopted. It performs interactive modeling of temperature and humidity features at two granularities: macroscopic semantics and microscopic numerical values, and stabilizes and enhances the robustness of co-coding through a perturbation correction mechanism.

[0033] First, the time-series feature autocorrelation matrix submodule 1331 is used to deeply mine the time-series features of a single input environmental parameter (i.e., temperature or humidity), and amplify its internal time-series dependence by constructing an autocorrelation matrix. Furthermore, the time-series feature autocorrelation matrix submodule 1331 can anchor the most critical core information affecting deformation through autocorrelation decomposition, thereby providing high-quality, denoised, and focused feature input for subsequent cross-parameter collaborative analysis.

[0034] Here, the input environmental temperature time-series correlation feature encoding vector is denoted as follows: The time-series associated feature encoding vector of ambient humidity, for example, denoted as respectively through feature mapping function (For example, a small multilayer perceptron network (MLP) or a kernel function) is projected onto a higher-dimensional feature space, and then its outer product is computed to obtain the semantic autocorrelation matrix. and This is to express and amplify the nonlinear autocorrelation within each time series feature.

[0035] In other words, the coupling effect of temperature and humidity on screen deformation in industrial scenarios is complex, exhibiting high-order nonlinearity and dynamic time-varying characteristics. The temporal features of ambient temperature and humidity contain a large amount of redundant noise, and direct processing would introduce interference signals. By constructing semantic autocorrelation matrices of the temporal correlation feature encoding vectors for ambient temperature and humidity, the autocorrelation relationships within the respective temporal features of temperature and humidity can be extracted. This helps to remove redundant information and lays the foundation for accurate subsequent analysis of the impact of temperature and humidity on screen deformation.

[0036] Furthermore, the generated semantic autocorrelation matrix... and The process involves decomposition and information purification to extract the core features that best characterize the effects of deformation. For example, the processing flow for ambient temperature is exactly the same as that for ambient humidity.

[0037] Specifically, the matrix Decomposed into multiple autocorrelation vectors by row. and for each Through a learnable weight matrix and bias Core information modulation is performed to obtain the modulation vector. And for each modulation vector, its core information anchoring factor is calculated. This factor is obtained through a nonlinear function (such as...) To measure the importance of this component: .

[0038] Furthermore, learnable anchor weights are used. For all autocorrelation vectors Perform a weighted summation, where the weights are determined by the corresponding anchoring factors. Regulation: .

[0039] Therefore, we can extract the environmental temperature time-series correlation feature encoding vector from the original vector, which may contain redundant information. Time-series associated feature encoding vector of ambient humidity In the process, core information anchoring encoding vectors strongly correlated with deformation are distilled out. and .

[0040] In other words, since the autocorrelation matrix may still contain some information with a relatively minor impact on screen deformation, core information anchoring can extract the most critical information that best reflects the relationship between temperature, humidity, and screen deformation. This allows focus on the main influencing factors and improves the accuracy of subsequent analysis. Specifically, this core information anchoring mechanism can extract core semantic anchors with strong deformation correlations (such as abrupt changes in temperature during rapid temperature increases or gradient features of continuous humidity accumulation) from two autocorrelation matrices, filtering out irrelevant noise (such as occasional anomalies from temperature sensors). Essentially, this process constructs a preprocessing link of "feature purification - semantic focus," ensuring that subsequent interactive modeling focuses only on key environmental disturbance patterns.

[0041] Then, the dual-granularity collaborative feature calculation and perturbation correction submodule 1332 is used to perform collaborative analysis, while other purified temperature and humidity core features, i.e., core information anchored encoding vectors, are received. and Furthermore, the synergistic effect of features is calculated in parallel from two dimensions: "feature granularity" (macro-semantic level) and "eigenvalue granularity" (micro-numerical level). Moreover, this application introduces a cross-granularity perturbation correction mechanism, which uses information from one granularity to stabilize and correct the representation of another granularity, thereby suppressing noise and enhancing the robustness of the model.

[0042] Specifically, the core information of environmental temperature time series features is anchored to the encoding vector. Anchoring encoding vector of core information of environmental humidity time series features First, by adding the two core information anchored encoding vectors, and then using activation functions (such as ReLU) and linear layers, their collaborative patterns in macroscopic semantics (e.g., the combined effect of rising temperature trends and stable humidity trends) are captured. This allows for the generation of a feature-granular deep co-encoding vector of environmental parameters and deformation effects through feature-granular deep co-encoding. : .

[0043] in and These are the weight matrix and the bias vector, respectively.

[0044] Simultaneously, the two core information anchored encoding vectors are multiplied element-wise to capture their non-linear dependencies on specific values ​​(for example, when the temperature exceeds 40°C, the impact of a 1% increase in humidity on deformation increases exponentially), so as to obtain the environmental parameter-deformation influence feature value granularity deep co-encoding vector through feature value granularity deep co-encoding. : .

[0045] Furthermore, since the information distribution of the two granularity encoding vectors may be inconsistent, it is necessary to utilize the relationship between them for mutual correction, i.e., to perform cross-granularity feature manifold perturbation correction, in order to obtain a more stable and accurate feature representation. Firstly, when calculating the corrected granularity encoding of the feature values, the difference response vector is calculated first. This reflects the difference between the two granularity coding methods, where the difference response vector... Each eigenvalue Deep co-encoding vector with environmental parameters - deformation influence features granularity Each eigenvalue And environmental parameters - deformation influence feature value granularity deep co-encoding vector Each eigenvalue The relationship is: .

[0046] Then, using this difference response vector To correct : .

[0047] Then, when calculating the corrected feature granularity encoding, firstly... Exponential modulation is performed to amplify the signal, i.e. and for the corrected Cosine modulation is performed to introduce nonlinear periodicity, i.e. Then, the two are multiplied positionally to obtain the corrected feature granularity encoding vector: .

[0048] Thus, we obtain the corrected environmental parameter-deformation influence feature granularity deep co-encoding vector. And the corrected environmental parameters - deformation influence feature value granularity deep co-encoding vector It not only contains coordinated information on temperature and humidity, but its internal characteristic manifold is also smoother and more stable, making it more resistant to interference such as sensor noise.

[0049] In other words, the effects of temperature and humidity on screen deformation are not independent but rather coupled. Deep co-coding at the feature granularity level, by anchoring the core information of environmental temperature time series features to the encoding vectors of environmental humidity time series features, can comprehensively consider the synergistic effects of temperature and humidity at the feature level, capturing the complex interactions between them, and thus more accurately characterizing the correlation between environmental parameters and screen deformation. Besides the synergistic effects at the feature level, there are also complex relationships between temperature and humidity feature values. By performing deep co-coding at the feature value granularity level, the synergistic effects between temperature and humidity feature values ​​can be further explored, analyzing the impact of environmental parameters on screen deformation from a more detailed perspective and supplementing information that may be missed in feature granularity encoding. That is, the synergistic effect of temperature and humidity on deformation is analyzed through dual-granularity interactive modeling, namely the response interaction between environmental temperature and environmental humidity (feature granularity and feature value granularity). It is worth mentioning that here, the environmental parameter-deformation influence feature granularity deep co-encoding vector can capture the coupling relationship between temperature and humidity at the macroscopic semantic level, such as the phase matching between temperature fluctuation cycle and humidity change trend on the overall impact of deformation; while the environmental parameter-deformation influence feature value granularity deep co-encoding vector can quantify the nonlinear dependence between temperature and humidity at the specific numerical level, such as the exponential jump in the contribution weight of humidity increment to deformation when the temperature value exceeds a threshold at a certain moment. The purpose of this operation is to establish a cross-parameter dynamic mapping model of temperature and humidity-deformation response, breaking through the limitations of single-parameter analysis or linear superposition.

[0050] Furthermore, in the analysis of the synergistic effect of temperature and humidity values ​​on deformation, when the temperature or humidity value exceeds the critical threshold within a certain time window, the corresponding humidity / temperature increment gradient will generate non-uniform diffusion perturbation at the eigenvalue interaction level. Therefore, through perturbation correction, the response lag caused by sudden temperature / humidity changes and the local manifold distortion caused by sensor noise can be effectively suppressed, ensuring that the corrected environmental parameter-deformation influence feature granularity depth co-encoding vector is well-defined. And the corrected environmental parameters - deformation influence feature value granularity deep co-encoding vector The final fusion generates a single, comprehensive environmental parameter-deformation influence deep co-encoding method. When describing complex patterns such as deformation accelerated by high temperature and humidity, the manifold topology maintains accurate feature representation in the parameter space. In other words, through a perturbation correction mechanism, it can capture both the macroscopic phase-matching features of temperature cycles and humidity trends, and accurately quantify nonlinear deformation transitions triggered by specific temperature and humidity combination thresholds.

[0051] Finally, the corrected co-coding feature fusion submodule 1333 combines the two high-quality co-coding vectors after dual-granularity modeling and perturbation correction, namely the corrected environmental parameter-deformation influence feature granularity depth co-coding vector. And the corrected environmental parameters - deformation influence feature value granularity deep co-encoding vector The final fusion is performed to generate a single, comprehensive environmental parameter-deformation effect deep co-encoding vector.

[0052] Specifically, a fusion function can be used here. To merge the two corrected encoded vectors, this function can be a simple concatenation followed by a wiring layer, or a more complex attention mechanism to dynamically determine the weights for extracting information from each vector: .

[0053] The environmental parameter-deformation influence deep co-coding vector This method comprehensively and accurately captures the nonlinear and dynamic coupling effects of temperature and humidity on screen deformation, providing a solid foundation for subsequent deformation coordinate compensation. Specifically, it achieves this by co-encoding a deep, granular vector of corrected environmental parameters and deformation influence features. And the corrected environmental parameters - deformation influence feature value granularity deep co-encoding vector Fusion enables the comprehensive and accurate representation of the impact of environmental parameters on screen deformation by utilizing information from both levels. Specifically, the model analyzes the generated environmental parameter-deformation influence co-encoding vector to identify key patterns of temperature and humidity interaction (such as abrupt changes in material expansion rate caused by the superposition of high temperature and high humidity), and quantifies the contribution ratio of the two at different time scales (e.g., temperature dominates short-term deformation, and humidity dominates long-term creep). This provides a feature interaction basis that combines accuracy and robustness for dynamic touch calibration.

[0054] Compared to traditional methods that typically fuse multimodal inputs at a single scale, the first-level subunit 133 of the deep co-coding of environmental parameters, through dual analysis at the feature granularity (macro semantics) and feature value granularity (micro numerical values), enables it to simultaneously capture the macro trend matching of temperature and humidity changes (such as the combination pattern of "gradual increase in temperature" and "sudden drop in humidity") and the precise numerical dependence at the micro level (such as "humidity influence factor x2 when temperature > 40℃"), greatly improving the ability to model complex nonlinear coupling relationships.

[0055] Furthermore, before engaging in multimodal interaction, the first-level subunit 133 of the deep co-coding of environmental parameters, through a core information anchoring mechanism based on self-correlation decomposition, first mines the temporal autocorrelation within each modality, and then anchors and extracts the most critical information for the final task (deformation prediction), which can effectively filter noise and make subsequent collaborative analysis more focused and efficient.

[0056] Furthermore, the first-level subunit 133 of the deep co-encoding of environmental parameters treats two co-encodings of different granularities as an interacting system through a cross-granularity feature manifold perturbation correction mechanism. It uses the output of one granularity to dynamically correct the other, and vice versa. Based on feedback correction in control theory, it effectively suppresses feature manifold distortion caused by sensor noise or sudden changes in environmental parameters, ensuring that the finally generated co-encoding vector maintains continuity and smoothness in the parameter space, thereby guaranteeing the robustness and generalization ability of the model.

[0057] Therefore, the first-level subunit 133 of the deep co-encoding of environmental parameters can solve the nonlinear coupling modeling problem that cannot accurately characterize the complex and nonlinear synergistic effect between temperature and humidity (such as the exponential growth of material deformation under the combined effect of high temperature and high humidity) when only considering the single influence of temperature and humidity or simple linear superposition. It also solves the problem of poor dynamic time-varying adaptability of static calibration, which cannot adapt to the dynamic changes of temperature and humidity and their cumulative effects, as well as the problem of characteristic noise interference in the original sensor data that will interfere with the accuracy of deformation prediction model.

[0058] In this embodiment, the deformation coordinate compensation amount generation first-level subunit 134 is used to input the environmental parameter-deformation influence depth co-encoding vector into an RNN-based sequence decoder to obtain the lateral deformation coordinate compensation amount and the longitudinal deformation coordinate compensation amount. It should be understood that the cumulative effect of temperature and humidity on screen deformation in industrial environments has a significant time dependence. For example, prolonged high temperatures not only cause instantaneous thermal expansion of materials but may also induce polymer layer creep (plastic deformation that gradually increases over time), while slow changes in humidity may continuously alter the stress distribution of the screen through moisture absorption over several hours. Therefore, the environmental parameter-deformation influence depth co-encoding vector is further input into an RNN-based sequence decoder to obtain the lateral deformation coordinate compensation amount and the longitudinal deformation coordinate compensation amount. Through the design of the RNN sequence decoder, the system aims to establish a dynamic evolution model of the deformation compensation amount. The environmental parameter-deformation influence deep co-encoding vector has already extracted the nonlinear features of the temperature and humidity synergy through preprocessing. The RNN decoder, through a hidden state propagation mechanism, associates the current co-encoding vector with the historical compensation generation process. For example, if the temperature has been rising continuously and the humidity fluctuations have been gentle over the past ten minutes, the RNN can automatically adjust the weights to strengthen the temperature-dominated rapid deformation component in the horizontal compensation calculation; conversely, if the humidity is detected to be entering an accelerated rising phase, the cumulative humidity effect weight in the vertical compensation is dynamically increased. This temporal modeling capability allows the compensation strategy to adapt to the evolution trend of environmental parameters, rather than relying solely on instantaneous states. In other words, the recurrent memory mechanism of the RNN endows it with the ability to model long-term temporal dependencies, incorporating the cumulative influence of historical environmental parameters on deformation into the prediction of the current compensation, thereby overcoming the limitations of static models.

[0059] In this embodiment, the touch coordinate correction unit 140 is used to correct the original touch coordinates based on the lateral deformation coordinate compensation amount and the longitudinal deformation coordinate compensation amount to obtain corrected touch coordinates. It should be understood that correcting the original touch coordinates based on the lateral deformation coordinate compensation amount and the longitudinal deformation coordinate compensation amount can significantly improve the touch accuracy of the robot display terminal. That is, the corrected touch coordinates more accurately reflect the user's actual touch position, reducing coordinate offset errors caused by environmental interference. This allows operators to more accurately select the required function or execute corresponding instructions when performing touch operations, avoiding misoperations caused by coordinate offsets and improving the accuracy and reliability of the operation.

[0060] The following is a detailed explanation of a specific implementation process for "correcting the original touch coordinates based on the lateral deformation coordinate compensation amount and the longitudinal deformation coordinate compensation amount to obtain the corrected touch coordinates": First, the robot's display terminal control system continuously monitors and receives the lateral and longitudinal deformation coordinate compensation values ​​generated by the dynamic touch calibration module, while simultaneously acquiring the determined original touch coordinates. This is the fundamental data source for the entire calibration process, and its accuracy and real-time performance are crucial. In practical applications, this data may be transmitted through a dedicated data transmission channel. To ensure stable data transmission, a series of data verification and error correction mechanisms are employed. For example, a checksum is added during data transmission. The receiving end uses the checksum to determine whether the data is complete and accurate. If an error is detected, a retransmission is requested promptly, thus ensuring data reliability. Simultaneously, to facilitate data management and processing, a dedicated data buffer is established in the system to temporarily store this data, enabling the control system to read and process it systematically.

[0061] Next, based on the hardware characteristics and software algorithm design of the display terminal, coordinate correction rules need to be pre-defined. Determining these rules requires comprehensive consideration of various factors. Typically, the most common correction method is addition, where the horizontal deformation coordinate compensation is directly added to the horizontal coordinate of the original touch coordinate, and the vertical deformation coordinate compensation is added to the vertical coordinate. This simple and direct method can meet basic correction requirements in most cases. However, in some special display terminal applications, such as professional image processing equipment with extremely high screen precision requirements, or devices with special screen shapes and display characteristics, more complex calculation rules may be needed. For example, weighted calculations based on specific coefficients, setting different weights for the horizontal and vertical compensation amounts according to the screen deformation characteristics of different areas, can achieve more accurate correction. These complex calculation rules require extensive experimentation and data analysis during the system development phase to determine the most suitable parameters and algorithms.

[0062] After determining the calibration rules, the control system begins performing coordinate calibration calculations. This process places strict requirements on both the accuracy and speed of the calculations. To ensure accuracy, the control system employs high-precision numerical calculation methods. For example, when handling decimal operations, it uses high-precision fixed-point or floating-point arithmetic to avoid rounding errors during the calculation process. Even small accumulated errors can lead to deviations in the final calibration result, affecting the accuracy of touch operations. Simultaneously, to meet real-time requirements, the system uses an efficient computing architecture or optimized algorithms. For instance, it utilizes parallel computing technology to distribute the calibration tasks of multiple original touch coordinates across multiple computing cores for simultaneous processing, significantly improving overall processing efficiency. Thus, when the user performs continuous touch operations, the system can respond quickly and complete coordinate calibration promptly, ensuring smooth touch operations.

[0063] After the coordinate correction calculation is completed, the corrected coordinates still need to undergo range checking and correction. This is because the display screen has physical boundaries, and the corrected coordinates are only valid within the effective display area of ​​the screen. The control system will carefully compare the corrected coordinates with the boundary coordinates of the display screen. If the corrected horizontal coordinate is less than the minimum horizontal coordinate of the display screen (usually 0), it means that the touch operation has exceeded the left boundary of the screen, and the system will correct the horizontal coordinate to 0; conversely, if it is greater than the maximum horizontal coordinate of the display screen, it will be corrected to the maximum horizontal coordinate. The same process is used for the vertical coordinate: if the vertical coordinate is less than the minimum vertical coordinate, it will be corrected to the minimum vertical coordinate; if it is greater than the maximum vertical coordinate, it will be corrected to the maximum vertical coordinate. Through this strict range checking and correction mechanism, it is ensured that the corrected touch coordinates are always within the effective range of the display screen, avoiding abnormal situations caused by touch operations exceeding the screen boundaries, such as accidentally triggering function buttons at the edge of the screen or the phenomenon of no touch response.

[0064] After coordinate range checking and correction, the corrected touch coordinates are considered accurate and valid. The control system outputs these corrected touch coordinates and transmits them to the subsequent processing modules of the display terminal. These modules perform corresponding operations based on the corrected touch coordinates, such as triggering the corresponding function button on the screen. When a user touches an icon on the screen, the system accurately identifies the icon's position based on the corrected coordinates and then executes the corresponding function. During the output process, to ensure the stability and accuracy of data transmission, a verification mechanism is used again to verify the transmitted corrected touch coordinate data, ensuring that no errors occur during data transmission, thereby guaranteeing the accuracy and reliability of the entire touch operation.

[0065] In summary, the dynamic touch calibration module 100 is clearly described. It employs artificial intelligence-based data processing and analysis technology. First, it determines the original touch coordinates based on the acquired original touch signal. Then, it separates the acquired temperature and humidity time sequence into independent time-series signals according to parameter dimensions. Next, it uses a one-dimensional convolutional network to extract the local temporal correlation features of temperature and humidity to capture dynamic fluctuation patterns such as temperature steps and gradual humidity changes. Subsequently, it fuses the temporal features of temperature and humidity through deep collaborative analysis of environmental parameters to analyze the nonlinear coupling effect of the two on screen deformation. Furthermore, it uses a sequence decoder to generate lateral and vertical deformation compensation amounts. Finally, the compensation amounts are superimposed on the original touch coordinates to form a closed-loop feedback correction. In this way, touch errors under environmental disturbances can be eliminated in real time.

[0066] In summary, the robot display terminal based on the embodiments of this application is explained. It consists of a display terminal and an adapter. The display terminal includes a display screen, function buttons, a safety switch, LED indicators, and an encoder knob. The adapter includes independently operating VGA and RS232 interfaces, used to receive video signals from the controller and to achieve data communication, respectively. The adapter also has a built-in dynamic touch calibration module, which can correct the touch coordinate offset of the display screen according to the ambient temperature and humidity. This effectively ensures the accuracy and reliability of touch operation of the robot display terminal in different environments, thereby improving the user experience and ensuring stable device performance.

[0067] Figure 4 This is a flowchart of an adaptive calibration method for a robot display terminal according to an embodiment of this application. Figure 4 As shown, the adaptive calibration method for a robot display terminal includes: S110, acquiring a time sequence of environmental parameters collected by a temperature and humidity sensor, wherein the environmental parameters include ambient temperature and ambient humidity; S120, acquiring an original touch signal and determining the original touch coordinates based on the mapping relationship between the original touch signal and theoretical coordinates; S130, performing sequence encoding and decoding on the time sequence of environmental parameters to obtain lateral deformation coordinate compensation and longitudinal deformation coordinate compensation; S140, correcting the original touch coordinates based on the lateral deformation coordinate compensation and the longitudinal deformation coordinate compensation to obtain corrected touch coordinates.

[0068] Here, those skilled in the art will understand that the specific operations of each step in the above-described adaptive calibration method for the robot display terminal have been referenced above. Figures 1 to 3 The dynamic touch calibration module of the robot display terminal is described in detail in the description, and therefore its repeated description will be omitted.

[0069] In summary, the adaptive calibration method for a robot display terminal based on the embodiments of this application is explained. It employs artificial intelligence-based data processing and analysis technology. First, it determines the original touch coordinates based on the acquired original touch signal. Then, it separates the acquired temperature and humidity time sequence into independent time-series signals according to parameter dimensions. Next, it uses a one-dimensional convolutional network to extract local temporal correlation features of temperature and humidity to capture dynamic fluctuation patterns such as temperature steps and gradual humidity changes. Subsequently, it uses deep collaborative analysis of environmental parameters to fuse the temporal features of temperature and humidity to analyze the nonlinear coupling effect of the two on screen deformation. Furthermore, it uses a sequence decoder to generate lateral and vertical deformation compensation amounts. Finally, the compensation amounts are superimposed on the original touch coordinates to form a closed-loop feedback correction. In this way, touch errors under environmental disturbances can be eliminated in real time.

[0070] In other embodiments of this application, a robot display terminal is also provided. Figure 5 This is a front structural diagram of a robot display terminal according to other embodiments of this application. Figure 6 This is a schematic diagram of the rear structure of a robot display terminal according to other embodiments of this application. Figure 5 and Figure 6 As shown, the three-way selector switch and emergency stop switch are core safety and mode control components. The three-way selector switch is used to switch the robot's working mode (such as manual, automatic, or standby) to adapt to different task scenarios; the emergency stop switch is used to immediately cut off the power or terminate the operation program in an emergency to ensure the safety of personnel and equipment. LED indicator lights are located in conspicuous positions and provide intuitive feedback on the real-time status of the system through color changes (such as solid green for normal operation and flashing red for faults), making it easy for users to quickly identify abnormalities. Function buttons provide basic interactive operations, such as inputting commands, selecting function options, and setting parameters, helping users to perform various operations and controls on the robot. The rotary encoder is typically used for precise parameter adjustment or browsing menu options. Rotation operation is used to increase or decrease values ​​or switch options, enabling continuous value adjustment or option switching. The three-state safety switch design emphasizes system safety and mode management. Its three states (such as "running," "maintenance," and "locked") correspond to different operating permissions or equipment operating conditions. For example, unauthorized operation is prohibited in the "locked" state, or technicians are allowed to perform debugging in the "maintenance" mode. USB sockets, as standard interfaces, are mainly used for connecting external storage devices, data transfer, or firmware upgrades. In some scenarios, they can also power external sensors or accessories. Their compatibility (such as USB 2.0 / 3.0) and protective design (such as dust cover) need to be determined based on the actual application requirements.

Claims

1. A robot display terminal, characterized in that, include: The display terminal includes a display screen, multiple function buttons, a safety switch, LED indicators, and an encoder knob; the adapter includes an interface module. The interface module includes a VGA interface and an RS232 interface. The VGA interface is used to receive video signals from the controller; the RS232 interface is used to realize data communication with the controller. The VGA interface and the RS232 interface are independent of each other.

2. The robot display terminal according to claim 1, characterized in that, The adapter also includes a dynamic touch calibration module, which is used to correct the touch coordinate offset of the display screen based on ambient temperature and humidity.

3. The robot display terminal according to claim 2, characterized in that, The dynamic touch calibration module includes: An environmental parameter data acquisition unit is used to acquire a time sequence of environmental parameters collected by a temperature and humidity sensor, the environmental parameters including ambient temperature and ambient humidity. The original touch coordinate determination unit is used to acquire the original touch signal and determine the original touch coordinates based on the mapping relationship between the original touch signal and the theoretical coordinates. The deformation coordinate compensation calculation unit is used to perform sequence encoding and decoding on the time queue of the environmental parameters to obtain the lateral deformation coordinate compensation and the longitudinal deformation coordinate compensation. A touch coordinate correction unit is used to correct the original touch coordinates based on the lateral deformation coordinate compensation amount and the longitudinal deformation coordinate compensation amount to obtain the corrected touch coordinates.

4. The robot display terminal according to claim 3, characterized in that, The deformation coordinate compensation calculation unit includes: The first-level subunit of the environmental parameter data division is used to divide the time queue of the environmental parameters according to the environmental parameter sample dimension to obtain the time queue of environmental temperature and the time queue of environmental humidity. The first-level subunit for environmental parameter temporal coding is used to perform one-dimensional convolution-based sequence coding on the time queue of environmental temperature and the time queue of environmental humidity to obtain the time-series associated feature coding vector of environmental temperature and the time-series associated feature coding vector of environmental humidity. The first-level sub-unit of deep co-coding of environmental parameters is used to input the environmental temperature time-series correlation feature encoding vector and the environmental humidity time-series correlation feature encoding vector into the deep co-analysis network of environmental parameters to obtain the deep co-coding vector of environmental parameters-deformation influence. A first-level sub-unit for generating deformation coordinate compensation is used to input the environmental parameter-deformation influence depth co-encoding vector into an RNN-based sequence decoder to obtain the lateral deformation coordinate compensation and the longitudinal deformation coordinate compensation.

5. The robot display terminal according to claim 4, characterized in that, The first-level subunit of the deep collaborative coding of environmental parameters includes: The time-series feature autocorrelation matrix submodule is used to perform time-series feature amplification and autocorrelation decomposition based on autocorrelation matrix time-series dependency amplification on the time-series features of the environmental temperature time-series correlation feature encoding vector and the environmental humidity time-series correlation feature encoding vector to obtain the core information anchoring encoding vector of the environmental temperature time-series feature and the core information anchoring encoding vector of the environmental humidity time-series feature. The dual-granularity collaborative feature calculation and perturbation correction submodule is used to perform parallel calculations of the synergistic effect of feature granularity and feature value granularity on the anchored encoding vector of the core information of the environmental temperature time series feature and the anchored encoding vector of the core information of the environmental humidity time series feature, and to perform perturbation correction to obtain the corrected environmental parameter-deformation influence feature granularity depth collaborative encoding vector and the corrected environmental parameter-deformation influence feature value granularity depth collaborative encoding vector; and, The corrected co-coding feature fusion submodule is used to fuse the corrected environmental parameter-deformation influence feature granularity depth co-coding vector and the corrected environmental parameter-deformation influence feature value granularity depth co-coding vector to obtain the environmental parameter-deformation influence depth co-coding vector.

6. The robot display terminal according to claim 5, characterized in that, The time-series feature autocorrelation matrix submodule is used for: The environmental temperature time-series correlation feature encoding vector and the environmental humidity time-series correlation feature encoding vector are respectively projected onto a high-dimensional feature space through a feature mapping function and their outer product is calculated to obtain the environmental temperature time-series semantic autocorrelation matrix and the environmental humidity time-series semantic autocorrelation matrix. After the environmental temperature time-series semantic autocorrelation matrix and the environmental humidity time-series semantic autocorrelation matrix are decomposed into multiple autocorrelation vectors by row, core information modulation is performed and core information anchoring factor is calculated. as well as The multiple autocorrelation vectors of the environmental temperature time-series semantic autocorrelation matrix and the environmental humidity time-series semantic autocorrelation matrix are weighted and summed with anchor weights and the core information anchoring factor to obtain the environmental temperature time-series feature core information anchoring encoding vector and the environmental humidity time-series feature core information anchoring encoding vector.

7. The robot display terminal according to claim 5, characterized in that, The dual-granularity collaborative feature calculation and perturbation correction submodule is used for: The anchored encoding vector of the core information of the environmental temperature time series feature and the anchored encoding vector of the core information of the environmental humidity time series feature are subjected to feature granularity deep co-coding to obtain the environmental parameter-deformation influence feature granularity deep co-coding vector; The anchored encoding vector of the core information of the environmental temperature time series feature and the anchored encoding vector of the core information of the environmental humidity time series feature are subjected to feature value granularity deep co-coding to obtain the environmental parameter-deformation influence feature value granularity deep co-coding vector; Cross-granularity feature manifold perturbation correction is performed on the environmental parameter-deformation influence feature granularity depth co-coding vector and the environmental parameter-deformation influence feature value granularity depth co-coding vector to obtain the corrected environmental parameter-deformation influence feature granularity depth co-coding vector and the corrected environmental parameter-deformation influence feature value granularity depth co-coding vector.

8. The robot display terminal according to claim 7, characterized in that, The cross-granularity characteristic manifold perturbation correction includes: Calculate the square of each feature value in the granular depth co-coding vector of the environmental parameter-deformation influence feature value, and subtract it from twice the corresponding feature value in the granular depth co-coding vector of the environmental parameter-deformation influence feature value to obtain the environmental parameter-deformation influence difference response representation vector; Calculate the square root of each feature value in the environmental parameter-deformation effect difference response representation vector, and divide it by the position of the environmental parameter-deformation effect feature granularity depth co-coding vector to obtain the corrected environmental parameter-deformation effect feature value granularity depth co-coding vector. Based on the environmental parameter-deformation influence feature granularity deep co-coding vector and the corrected environmental parameter-deformation influence feature value granularity deep co-coding vector, exponential modulation and cosine modulation are performed followed by positional dot product to obtain the corrected environmental parameter-deformation influence feature granularity deep co-coding vector.

9. An adaptive calibration method for a robot display terminal, characterized in that, include: A time sequence of environmental parameters collected by a temperature and humidity sensor is obtained, the environmental parameters including ambient temperature and ambient humidity; Acquire the original touch signal and determine the original touch coordinates based on the mapping relationship between the original touch signal and the theoretical coordinates; The time sequence of the environmental parameters is encoded and decoded to obtain the lateral deformation coordinate compensation amount and the longitudinal deformation coordinate compensation amount; Based on the lateral deformation coordinate compensation amount and the longitudinal deformation coordinate compensation amount, the original touch coordinates are corrected to obtain the corrected touch coordinates.

Images