Search

CN-122018693-A - Electronic skin sensing system and method for bionic interaction of animals

CN122018693ACN 122018693 ACN122018693 ACN 122018693ACN-122018693-A

Abstract

The invention relates to the field of electronic skin sensing, and discloses an electronic skin sensing system and method for bionic interaction of animals, wherein the electronic skin sensing system comprises the steps of collecting touch signals with different contact intensities, directions and vibration spectrums to form a basic feedback database; the method comprises the steps of mapping a real-time collected haptic signal vector to a basic feedback database through a matching algorithm, quickly determining candidate feedback situations according to a strategy rule, preprocessing recorded data under a preset trigger condition on the basis of executing the candidate feedback situations, uploading the preprocessed recorded data to a central processing platform to generate a new feedback situation, merging the new feedback situation into the basic feedback database, combining historical user data and current environmental conditions to construct a front feedback situation set and a rear feedback situation set, executing recording according to recently collected haptic sensation, and executing multi-modal verification on user behavior and physiological response data to generate an updated front feedback situation strategy. The invention has the advantages of improving interaction precision and efficiency.

Inventors

  • WANG DI
  • MENG QINGYU
  • XIAO YANG

Assignees

  • 深圳大深传感科技有限公司

Dates

Publication Date
20260512
Application Date
20260203

Claims (10)

  1. 1. An electronic skin sensing method for bionic interaction of animals is characterized by comprising the following steps: the method comprises the steps of collecting touch signals with different contact intensities, directions and vibration spectrums, and encoding the signals into indexable feedback situation sets through off-line batch processing to form a basic feedback database; Mapping the haptic signal vector acquired in real time to a basic feedback database through a matching algorithm, carrying out multi-dimensional interval judgment by combining signal amplitude, frequency spectrum characteristics and environmental weights, and rapidly determining candidate feedback situations according to strategy rules; recording current tactile signal characteristics, an execution mode, user behaviors and emotion changes on the basis of executing candidate feedback situations, preprocessing recorded data under a preset trigger condition, and uploading the preprocessed recorded data to a central processing platform to generate a new feedback situation; integrating the new feedback situation into a basic feedback library, and combining historical use data of a user with current environmental conditions, counting the use probability and success rate of different feedback situations, and constructing a front feedback situation set and a rear feedback situation set; And (3) collecting the front feedback situation and the rear feedback situation, performing multi-mode verification according to the recently acquired touch executing record and the user behavior and physiological response data, continuously optimizing and fine-tuning the executing strategy, and generating an updated front feedback situation strategy.
  2. 2. The method for electronic skin sensing for bionic interaction of animals according to claim 1, wherein the process of forming the basic feedback database is: A multi-channel piezoelectric and resistance sensing unit is arranged on the surface of the electronic skin, and multi-dimensional characteristics of contact intensity, direction, vibration spectrum, contact area and friction coefficient in a touch signal are collected; Performing time synchronization, anomaly rejection, frequency spectrum filtering and amplitude normalization on the acquired signals; And mapping the processed haptic signals to an index space capable of being quickly searched by adopting a high-dimensional vector coding algorithm and a local sensitive hash index to generate a standardized and multi-mode basic feedback database for real-time haptic signal matching, candidate feedback situation generation and abnormal situation identification.
  3. 3. The electronic skin sensing method for bionic interaction of animals according to claim 2, wherein the process of mapping the real-time collected haptic signal vector to the basic feedback database through the matching algorithm is as follows: performing multi-dimensional feature extraction such as amplitude distribution, frequency spectrum composition, vibration mode, time sequence change and the like on the touch signal vector acquired in real time to generate a complete real-time feature vector; Comparing the real-time feature vector with the feature vector of each feedback situation in the basic feedback database, calculating the similarity through dimension-by-dimension feature matching, identifying the feedback situation closest to the current signal, and generating a candidate feedback situation set; And sequencing the situations in the candidate set according to the matching degree and the applicability by combining the haptic acquisition environment information, the user behavior preference and the context condition of the current interaction task, and outputting a candidate feedback situation subset executed with priority.
  4. 4. The electronic skin sensing method for bionic interaction of animals according to claim 3, wherein the multi-dimensional interval judgment process by combining the signal amplitude, the frequency spectrum characteristic and the environment weight is as follows: Carrying out dimension-wise comparison on the candidate feedback situation subsets and the real-time tactile signal vectors, wherein the dimension-wise comparison comprises matching degree of amplitude intervals, frequency spectrum similarity, correlation of vibration modes and time offset alignment; introducing a multidimensional environment weighting factor based on the environmental illumination, temperature, material reflection characteristics and tactile surface friction characteristics, and normalizing and punishing and correcting the matching degree; And screening out feedback situations with all matching degrees exceeding the self-adaptive threshold by using a threshold judgment rule and a multi-layer confidence fusion mechanism, and generating a weighted candidate feedback situation set.
  5. 5. The electronic skin sensing method for bionic interaction of animals according to claim 4, wherein the process of rapidly determining candidate feedback situations according to policy rules is as follows: Calculating a candidate feedback situation priority score by adopting a multi-objective optimization algorithm based on the weighted candidate feedback situation set and combining the user behavior history, the current task target and the safety constraint; And (3) invoking a strategy rule base, and screening candidate situations with high scores by combining the power consumption limit, response delay and execution safety boundary to generate a final candidate feedback situation.
  6. 6. The method for electronic skin perception of animal bionic interaction according to claim 5, wherein the recording of current tactile signal characteristics, execution patterns and user behavior and mood changes is: the final candidate feedback situation execution data including contact pressure, vibration intensity, tangential force and response time are collected in parallel through an electronic skin controller and an embedded sensor; Collecting user behavior and physiological signals in real time, including hand movement track, myoelectricity, heart rate, galvanic skin response and facial expression indexes, and storing the signals and tactile sensation execution data synchronously to form a multi-mode execution record; Binding the multi-mode execution record with the candidate feedback situation list to generate a record data packet.
  7. 7. The electronic skin sensing method for bionic interaction of animals according to claim 6, wherein the process of generating a new feedback situation is: when the recorded data packet reaches a preset time window, carrying out abnormal rejection, normalization and feature enhancement on the touch executing data, the user behavior data and the physiological response signals contained in the data packet to obtain a clean feature vector; transmitting the clean feature vector to a central processing platform in an encryption mode; And comparing and analyzing the uploaded data by combining the historical feedback situation, the historical execution record and the behavior trend to generate a new feedback situation.
  8. 8. The electronic skin sensing method for bionic interaction of animals according to claim 7, wherein the process of constructing the front and rear feedback situation sets is: summarizing the new feedback situation and all the historical feedback situations in the basic feedback library to form a complete feedback situation set; multidimensional comparison analysis is carried out on each feedback situation in the set, wherein the multidimensional comparison analysis comprises a tactile signal amplitude, a frequency spectrum composition, a vibration mode and a time sequence change characteristic; The response effect of each situation under different interaction scenes is counted by combining historical use data of the user, wherein the response effect comprises user response time, satisfaction degree score, reuse frequency and physiological signal consistency, and situation performance score is generated; Distributing dynamic weights to each feedback situation according to the current environmental conditions, user behavior preference and task context; And dividing feedback situations in the whole set according to situation performance scores and environment weights, and constructing a front feedback situation set and a rear feedback situation set.
  9. 9. The electronic skin sensing method for bionic interaction of animals according to claim 8, wherein the process of generating the updated pre-feedback situation strategy is: The front feedback situation set and the rear feedback situation set are subjected to feature alignment and trend analysis with recently acquired touch executing records, user behavior data and physiological response signals; Based on the alignment and analysis results, comprehensively scoring the execution sequence, the touch strength, the frequency spectrum parameters and the triggering conditions of the feedback situation by utilizing a multi-factor weighted evaluation method, and evaluating the applicability and the priority of each situation under different environmental conditions and user behaviors; fine tuning low scoring situations including adjusting haptic intensity, spectral parameters, execution delay, and order rearrangement; And generating an updated preposed feedback situation strategy by combining the triggering relation and the dependent conditions between the preposed situation and the postposed situation.
  10. 10. An electronic skin-aware system for biomimetic interaction of animals, applied to a method according to any one of claims 1-9, comprising: The data acquisition module acquires touch signals with different contact intensities, directions and vibration spectrums from the multi-channel touch sensing unit on the surface of the electronic skin, and performs batch processing to form a basic feedback database; The signal matching module is used for carrying out multidimensional feature comparison and environment weighting judgment on the haptic signal vector acquired in real time and the situation vector in the basic feedback database to generate a priority candidate feedback situation; the execution recording module is used for collecting the touch signal, the user behavior and the emotion change of the execution situation in real time, processing and uploading the record under the triggering condition and generating a new feedback situation; the situation construction module combines the new feedback situation with the historical feedback database, counts the use probability and the success rate, and constructs a front feedback situation set and a rear feedback situation set; and the strategy optimization module is used for verifying and fine-tuning the execution sequence, the touch strength and the frequency spectrum parameters based on the front feedback situation set, the rear feedback situation set and the recent multi-mode execution record, and generating an updated front feedback situation strategy.

Description

Electronic skin sensing system and method for bionic interaction of animals Technical Field The invention relates to the field of electronic skin sensing, in particular to an electronic skin sensing system and method for bionic interaction of animals. Background With the development of bionic robots and intelligent interaction technologies, electronic skin has important application value in animal bionic interaction, multi-dimensional tactile information such as pressure, vibration and tangential force is acquired through a multi-channel tactile sensor, and the sensing capability is provided for an interaction system, however, the existing electronic skin still has obvious defects in practical application, a single tactile signal processing mode is lack of comprehensive analysis on amplitude, frequency spectrum, vibration mode and time sequence characteristics, so that tactile feedback precision is low, response delay is caused, real-time signal and historical feedback matching lacks of dynamic weighting of environmental conditions and user behavior preference, feedback selection is not self-adaptive enough, complete records on tactile output and user behavior and physiological response are lacking in an execution process, a data accumulation and strategy optimization mechanism cannot be formed, generation of a new tactile situation and database update are dependent on manual or fixed rules, automatic, multi-mode processing and dynamic priority management are lacked, and individuation and long-term optimization are difficult to realize. Disclosure of Invention Aiming at the defects of the prior art, the invention provides an electronic skin sensing system and method for animal bionic interaction, which have the advantages of improving interaction precision and efficiency and solve the problems in the background art. In order to achieve the aim of improving interaction precision and efficiency, the invention provides the following technical scheme that the electronic skin sensing method for animal bionic interaction comprises the following steps: the method comprises the steps of collecting touch signals with different contact intensities, directions and vibration spectrums, and encoding the signals into indexable feedback situation sets through off-line batch processing to form a basic feedback database; Mapping the haptic signal vector acquired in real time to a basic feedback database through a matching algorithm, carrying out multi-dimensional interval judgment by combining signal amplitude, frequency spectrum characteristics and environmental weights, and rapidly determining candidate feedback situations according to strategy rules; recording current tactile signal characteristics, an execution mode, user behaviors and emotion changes on the basis of executing candidate feedback situations, preprocessing recorded data under a preset trigger condition, and uploading the preprocessed recorded data to a central processing platform to generate a new feedback situation; integrating the new feedback situation into a basic feedback library, and combining historical use data of a user with current environmental conditions, counting the use probability and success rate of different feedback situations, and constructing a front feedback situation set and a rear feedback situation set; And (3) collecting the front feedback situation and the rear feedback situation, performing multi-mode verification according to the recently acquired touch executing record and the user behavior and physiological response data, continuously optimizing and fine-tuning the executing strategy, and generating an updated front feedback situation strategy. Preferably, the process of forming the base feedback database is: A multi-channel piezoelectric and resistance sensing unit is arranged on the surface of the electronic skin, and multi-dimensional characteristics of contact intensity, direction, vibration spectrum, contact area and friction coefficient in a touch signal are collected; Performing time synchronization, anomaly rejection, frequency spectrum filtering and amplitude normalization on the acquired signals; And mapping the processed haptic signals to an index space capable of being quickly searched by adopting a high-dimensional vector coding algorithm and a local sensitive hash index to generate a standardized and multi-mode basic feedback database for real-time haptic signal matching, candidate feedback situation generation and abnormal situation identification. Preferably, the process of mapping the haptic signal vector acquired in real time to the basic feedback database through the matching algorithm is as follows: performing multi-dimensional feature extraction such as amplitude distribution, frequency spectrum composition, vibration mode, time sequence change and the like on the touch signal vector acquired in real time to generate a complete real-time feature vector; Comparing the real-time feature vector with the feat