Infrared alignment signal data processing method applied to robot

Abstract

This invention discloses an infrared alignment signal data processing method for robots, relating to the field of robot control technology. The method includes: analyzing the changing trend of infrared signal intensity based on the infrared signal evolution sequence and determining the heading adjustment state to obtain the heading control degrees of freedom; constructing a spatiotemporal frequency coherence tensor of the infrared alignment signal based on the heading control degrees of freedom, and using a geometric propagation algorithm to identify and separate non-target interference components to obtain continuous signal variation segments; mapping the infrared signal intensity to the logarithmic domain based on the continuous signal variation segments, and constructing an offset decoupling equation to obtain the offset direction and offset degree; mapping the offset direction and offset degree to heading correction and position correction amounts, and performing closed-loop correction to generate alignment control commands. This invention improves the continuity and reliability of industrial control software signals in complex environments by constructing a spatiotemporal frequency coherence tensor and using a geometric propagation algorithm to separate non-target interference components.

Description

Technical Field

[0001] This invention relates to the field of robot control technology, and in particular to an infrared alignment signal data processing method for robots. Background Technology

[0002] With the widespread application of mobile robots, autonomous mobile platforms, and service robots in industrial manufacturing, smart warehousing, automatic charging, and unmanned inspection, environmentally-based autonomous alignment and precise docking technologies have gradually become an important research direction in the field of robot motion control. Among various sensing methods, infrared alignment technology is widely used in robot charging docking, station docking, and close-range attitude calibration scenarios due to its advantages such as simple structure, fast response speed, strong resistance to visible light interference, and low cost. Existing infrared alignment typically involves collecting the infrared signal intensity of each channel in industrial control software and determining the relative positional deviation between the robot and the target based on the intensity magnitude and difference, thereby driving the robot to perform heading correction or position adjustment.

[0003] Existing infrared alignment technologies generally suffer from two shortcomings in data processing: First, they focus on instantaneous signal strength comparison and fixed threshold judgment, lacking analysis of the co-evolution of multi-channel infrared signals over time. This leads to jitter or false triggering in heading adjustment judgments when interference or signal abrupt changes occur, thus affecting the continuity and stability of robot motion. Second, existing technologies typically separate signal processing from motion control, failing to establish a dynamic mapping relationship between the structural characteristics of infrared signals and the degrees of freedom of heading control. This makes it difficult to achieve precise discrimination and constraint control of heading adjustment states, resulting in problems such as overcorrection, oscillation, or decreased alignment efficiency in complex scenarios. Summary of the Invention

[0004] In view of the aforementioned existing problems, the present invention is proposed.

[0005] Therefore, the present invention provides an infrared alignment signal data processing method for robots to solve the problems of unstable heading adjustment and difficulty in achieving fine control during infrared alignment.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: This invention provides an infrared alignment signal data processing method for robots, comprising: acquiring the infrared signal intensity of infrared alignment signals from multiple channels and performing time alignment to obtain an infrared signal evolution sequence; analyzing the changing trend of infrared signal intensity based on the infrared signal evolution sequence and determining the heading adjustment state to obtain the heading control degrees of freedom; constructing a spatiotemporal frequency coherence tensor of the infrared alignment signal based on the heading control degrees of freedom, and using a geometric propagation algorithm to identify and separate non-target interference components to obtain continuous signal variation segments; mapping the infrared signal intensity to the logarithmic domain based on the continuous signal variation segments, and constructing an offset decoupling equation to obtain the offset direction and offset degree; mapping the offset direction and offset degree to heading correction and position correction, and performing closed-loop correction to generate alignment control commands.

[0007] As a preferred embodiment of the infrared alignment signal data processing method for robots described in this invention, the infrared signal recording set includes channel identifier, acquisition time, and infrared signal intensity. The infrared signal intensity of the infrared alignment signals from multiple channels is collected, and time alignment is performed to obtain the infrared signal evolution sequence. The specific steps are as follows: The infrared signal intensity of infrared alignment signals from multiple channels is collected, and the corresponding collection time is recorded to obtain an infrared signal record set. Align the channel identifier with the acquisition time and map the infrared signal intensity to the same time base to generate a multi-channel infrared signal sequence.

[0008] As a preferred embodiment of the infrared alignment signal data processing method for robots described in this invention, the infrared signal evolution sequence is obtained by sequentially arranging the infrared signal intensity in ascending order according to the time sequence of the multi-channel infrared signal sequence.

[0009] As a preferred embodiment of the infrared alignment signal data processing method for robots described in this invention, the specific steps for analyzing the changing trend of infrared signal intensity based on the infrared signal evolution sequence are as follows: Based on the infrared signal evolution sequence, abrupt segments are marked and removed using the sliding window decomposition method to obtain a set of continuous and effective signal segments; Calculate the multi-channel signal consistency score within a set of continuous valid signal segments, perform spatiotemporal frequency coherence assessment, and obtain coherence labels; The same coherence markers are merged into coherent segments and arranged in chronological order to generate a coherent segment sequence.

[0010] In a preferred embodiment of the infrared alignment signal data processing method for robots described in this invention, the heading control degrees of freedom are obtained by comparing the continuity and consistency of coherent segment sequences, generating a heading adjustment state segment sequence, and performing adjustment state mapping.

[0011] As a preferred embodiment of the infrared alignment signal data processing method for robots described in this invention, the specific steps for constructing the spatiotemporal frequency coherence tensor of the infrared alignment signal based on the heading control degrees of freedom are as follows: The infrared signal evolution sequence is resampled and rearranged in a directionally restricted manner according to the heading control degree of freedom to generate a restricted infrared signal sequence. A spatial adjacency constraint matrix is ​​constructed based on a constrained infrared signal sequence. A unified representation of the co-evolution is generated by using a spatiotemporal-frequency joint embedding method.

[0012] In a preferred embodiment of the infrared alignment signal data processing method for robots described in this invention, the specific steps for obtaining the continuous signal variation segment are as follows: Geometric propagation is performed on the spatiotemporal frequency coherence tensor, and propagation consistency analysis is conducted to obtain non-target interference components; Non-target interference components in the spatiotemporal frequency coherence tensor are separated and integrated with temporal continuity constraints to obtain continuous signal variation segments.

[0013] As a preferred embodiment of the infrared alignment signal data processing method for robots described in this invention, the steps of mapping the infrared signal intensity to the logarithmic domain based on continuous signal variation segments and constructing an offset decoupling equation to obtain the offset direction and offset degree are as follows: Logarithmic domain mapping is performed on the intensity of multi-channel infrared signals within a continuous signal variation segment to generate a logarithmic domain infrared signal segment sequence. Calculate the logarithmic difference of the intensity of multi-channel infrared signals in the sequence of logarithmic domain infrared signals, and construct the multi-channel intensity gradient vector; Based on the multi-channel intensity gradient vector, the offset decoupling equation is constructed by odd-even grouping and linear combination according to the channel identifier, and decomposed into lateral offset component and heading offset component. The lateral offset component and the heading offset component are integrated over time and their signs are determined within the continuous signal variation segment to obtain the offset direction and offset degree.

[0014] As a preferred embodiment of the infrared alignment signal data processing method for robots described in this invention, the step of mapping the offset direction and offset degree to heading correction and position correction refers to using a direction amplitude coupling mapping algorithm to generate heading correction and position correction by using the offset direction as a symbol constraint and the offset degree as an amplitude scale.

[0015] In a preferred embodiment of the infrared alignment signal data processing method for robots described in this invention, the specific steps for generating alignment control commands are as follows: Based on the heading correction and position correction, a stable correction pair is generated by performing coupled constraint processing through an interlocking vibration suppression adjustment algorithm. The stable correction pairs are recursively updated and boundary limiting is performed to generate alignment control commands.

[0016] The beneficial effects of this invention are as follows: by constructing a spatiotemporal frequency coherence tensor and using a geometric propagation algorithm to separate non-target interference components, the co-evolution analysis of multi-channel infrared signals in the time, space, and frequency dimensions is realized, thereby improving the continuity and reliability of industrial control software in complex environments; by using logarithmic domain mapping and offset decoupling equations, the coupled signal is decomposed into lateral and heading offset components, realizing the accurate conversion of signal structural features into control quantities, and improving the stability and convergence of the robot's autonomous alignment process. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a flowchart of an infrared alignment signal data processing method applied to robots.

[0019] Figure 2 A flowchart for generating an infrared signal evolution sequence.

[0020] Figure 3 A flowchart for obtaining the heading control degrees of freedom.

[0021] Figure 4 A flowchart for generating alignment control commands.

[0022] Figure 5 This is a diagram showing the effect of coherent interference separation.

[0023] Figure 6 This is the convergence plot for offset decoupling. Detailed Implementation

[0024] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0025] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0026] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0027] Reference Figures 1-6 As one embodiment of the present invention, this embodiment provides an infrared alignment signal data processing method for robots, comprising the following steps: S1: Collect the infrared signal intensity of infrared alignment signals from multiple channels, perform time alignment, and obtain the infrared signal evolution sequence.

[0028] S1.1: Collect the infrared signal intensity of infrared alignment signals from multiple channels, record the corresponding acquisition time, and obtain an infrared signal record set.

[0029] Specifically, an infrared receiver is used to receive infrared alignment signals from multiple channels and output corresponding analog voltage signals and mark the corresponding channels to obtain channel identifiers; the analog voltage signals output from each channel of the infrared alignment signal are sampled sequentially to keep the analog voltage signals stable in each sampling period, and an analog-to-digital converter is used to convert the analog voltage signals to digital to obtain the infrared signal intensity corresponding to the infrared radiation intensity.

[0030] The current timing value of the timing circuit is read as the acquisition time, and the channel identifier, acquisition time and corresponding infrared signal intensity value are matched one by one to form an infrared signal record. The infrared signal records of all infrared alignment signal channels are summarized to obtain an infrared signal record set (including channel identifier, acquisition time and infrared signal intensity).

[0031] S1.2: Align the channel identifier with the acquisition time and map the infrared signal intensity to the same time reference to generate a multi-channel infrared signal sequence.

[0032] Specifically, the infrared signal records in the infrared signal record set are grouped according to the channel identifier, so that each group of infrared signal records corresponds to an infrared alignment signal channel; within each group of infrared signal records, they are sorted in ascending order according to the acquisition time to obtain the channel infrared signal record sequence arranged by acquisition time.

[0033] Extract the acquisition times appearing in the infrared signal recording sequence of all channels to construct a unified time reference sequence; for each acquisition time in the unified time reference sequence, search for infrared signal records with the same acquisition time in the infrared signal recording sequence of each channel; when a corresponding acquisition time exists, directly extract the corresponding infrared signal intensity; when no corresponding acquisition time exists, perform linear interpolation between two adjacent acquisition times to obtain the infrared signal intensity corresponding to the current acquisition time.

[0034] Following the time sequence of a unified time base sequence, the multi-channel infrared signal intensities corresponding to each acquisition moment are arranged in order of channel identifiers to generate a multi-channel infrared signal sequence.

[0035] S1.3: Based on the multi-channel infrared signal sequence, the infrared signal intensities are arranged in ascending order according to time sequence to obtain the infrared signal evolution sequence.

[0036] Specifically, the process involves traversing the unified time reference sequence in the multi-channel infrared signal sequence, sequentially reading the corresponding multi-channel infrared signal intensity at each acquisition time in the unified time reference sequence, while maintaining the channel identifier order unchanged.

[0037] The intensity of multi-channel infrared signals corresponding to the same acquisition time is combined according to the order of channel identifiers to form a time node data grid; the time node data grids of all acquisition times are counted and arranged in ascending order of time to obtain the infrared signal evolution sequence.

[0038] S2: Based on the infrared signal evolution sequence, analyze the changing trend of infrared signal intensity, determine the heading adjustment state, and obtain the heading control degrees of freedom.

[0039] S2.1: Based on the infrared signal evolution sequence, abrupt segments are marked and removed using the sliding window decomposition method to obtain a set of continuous and effective signal segments.

[0040] Specifically, the infrared signal evolution sequence is traversed in chronological order. Based on the channel identifier, the changes in infrared signal intensity corresponding to adjacent acquisition times in the infrared signal evolution sequence are statistically analyzed, and the changes in infrared signal intensity of multiple channels are arranged in chronological order to form an infrared signal intensity change sequence.

[0041] The infrared signal intensity change sequence is divided into multiple time windows of different lengths. Each time window covers several consecutive acquisition moments and is traversed sequentially by sliding one acquisition moment forward. Within each time window, the sign of the changes in the intensity of the multi-channel infrared signal is statistically analyzed, and the number of times the sign of the changes in the intensity of the multi-channel infrared signal is reversed within the time window is recorded.

[0042] The amplitude variation range of the multi-channel infrared signal intensity change within the time window is statistically analyzed. When the number of sign reversals within the time window is greater than the number of sign reversals within the adjacent time window, and the amplitude variation range is greater than the amplitude variation range within the adjacent time window, the acquisition time segment covered by the corresponding time window is marked as a sudden change segment.

[0043] The acquisition time segments marked as mutation segments are removed from the infrared signal evolution sequence, and the acquisition time segments that are not marked as mutation segments and are continuous in time are combined in chronological order to obtain a set of continuous and valid signal segments.

[0044] It should be noted that the sliding window decomposition method is an analysis method that segments a signal sequence along a time sequence. By dividing the time window and statistically comparing the signal change characteristics within the window coverage area, and updating the local feature description after each window slide (such as the number of sign reversals of the multi-channel infrared signal intensity changes, amplitude change range, average change rate, change rate variance, and the consistency ratio of change direction between channels), the overall signal is decomposed into multiple segments with local consistency or local anomaly characteristics, which are used to identify abrupt change segments, trend change segments, or continuous stable segments.

[0045] S2.2: Calculate the multi-channel signal consistency score within the set of continuous valid signal segments, perform spatiotemporal frequency coherence assessment, and obtain coherence labels.

[0046] Specifically, the algorithm iterates through each acquisition time in the set of continuous valid signal segments, extracts the intensity changes of the multi-channel infrared signal at the corresponding acquisition time, statistically analyzes the signs of the intensity changes of each channel's infrared signal within the same acquisition time, and calculates the multi-channel signal consistency score, expressed as:

[0047] ; in, For multi-channel signal consistency scoring, The total number of channel identifiers. For symbolic functions, For the first The change in infrared signal intensity on each channel identifier For channel identifier index.

[0048] It should be noted that by summing and normalizing the signs of the changes in the intensity of the multi-channel infrared signals, the degree of consistency in the direction of change is converted into a dimensionless proportional value. Therefore, the expression for calculating the consistency score of the multi-channel signals has a unified dimension.

[0049] The consistency scores of the multi-channel signals arranged in chronological order form a consistency score sequence. Within the set of continuous valid signal segments, the discrete Fourier transform is performed on the intensity change of the multi-channel infrared signal corresponding to each acquisition time to obtain the frequency component amplitude at the corresponding acquisition time, and then arranged in chronological order to form a frequency component amplitude sequence.

[0050] The expression for calculating the amplitude of the frequency component is:

[0051] ; in, The amplitude of the frequency component. The frequency component number. For the time of data collection In the The change in infrared signal intensity on each channel.

[0052] It should be noted that the infrared signal intensity involved in the calculation in the formula is in the dimension of voltage, and the trigonometric function operation itself does not introduce new physical quantity properties. Therefore, the calculated frequency component amplitude and the change in infrared signal intensity involved in the calculation have the same dimension, and the overall dimension remains consistent.

[0053] The consistency score sequence and the frequency component amplitude sequence are compared moment by moment to determine whether the changes in the time dimension are synchronized. When the changes are in the same direction for multiple consecutive acquisition moments, the corresponding acquisition moment is marked as a coherent state. When the direction is reversed, the corresponding acquisition moment is marked as an incoherent state, thus obtaining the coherence label.

[0054] S2.3: Merge identical coherence markers into coherent segments and arrange them in chronological order to generate a coherent segment sequence.

[0055] Specifically, the coherence markers are traversed in time order to record the coherence state (including coherent and incoherent states) corresponding to each acquisition time. When the coherence markers corresponding to multiple consecutive acquisition times remain in the same state, the start and end acquisition times of the consecutive coherence markers are recorded, and the corresponding time interval is taken as a coherence segment.

[0056] When the coherence marker changes state between adjacent acquisition times, the boundary recording of the current coherence segment ends and the recording of a new coherence segment boundary begins; all recorded coherence segments are arranged in ascending order of their starting acquisition times to generate a coherence segment sequence.

[0057] S2.4: Compare the persistence and consistency of the coherent segment sequence, generate the heading adjustment state segment sequence, and perform adjustment state mapping to obtain the heading control degrees of freedom.

[0058] Specifically, the duration of the coherent segment is obtained by counting the time difference between the start and end times of the coherent segment; the number of coherence markers within the coherent segment is counted, and the consistency result of the coherent segment is determined based on whether the coherence markers within the coherent segment remain consistent throughout the entire coherent segment.

[0059] Each coherent segment in the coherent segment sequence is compared segment by segment. When the duration of a coherent segment is greater than that of the adjacent coherent segment and the coherence marker within the coherent segment remains consistent throughout the segment, the corresponding coherent segment is marked as an effective heading adjustment segment. When the duration of a coherent segment is less than that of the adjacent coherent segment or the coherence marker within the coherent segment changes within the segment, the corresponding coherent segment is marked as a heading adjustment restriction segment, thus generating a heading adjustment state segment sequence.

[0060] Based on the temporal arrangement of the effective and restricted heading segments in the heading heading state segment sequence, the heading state mapping is performed. When the effective heading segment appears at the current time position, it is mapped to allow heading adjustment; when the restricted heading segment appears at the current time position, it is mapped to restrict heading adjustment, thus obtaining the heading control degrees of freedom.

[0061] like Figure 5As shown, the coherence index of the unseparated non-target interference components is compared with the coherence index after interference separation based on the spatiotemporal frequency coherence tensor combined with the geometric propagation algorithm. It can be seen that in the initial stage and within several interference enhancement intervals, the coherence index of the unseparated case fluctuates greatly and declines within local time windows, indicating that the coordinated consistency of the multi-channel infrared signal in the time, space, and frequency dimensions is weakened by the non-target interference. After adopting the geometric propagation separation algorithm, the coherence index remains in a relatively high range overall, with reduced fluctuation. A magnified view further shows that the difference is most significant in the interference concentration segment, and the separated curve maintains a continuous upward trend without abrupt instability. The results indicate that this invention, by constructing a spatiotemporal frequency coherence tensor and performing propagation path constraint separation of non-target interference components, effectively enhances the coordinated evolution capability of multi-channel infrared signals in complex environments, thereby improving the stability and reliability of industrial control software during continuous alignment processes.

[0062] S3: Based on the heading control degrees of freedom, construct the spatiotemporal frequency coherence tensor of the infrared alignment signal, and use the geometric propagation algorithm to identify and separate non-target interference components to obtain the continuous signal variation segment.

[0063] S3.1: Perform directional-restricted resampling and channel rearrangement on the infrared signal evolution sequence according to the heading control degrees of freedom to generate a restricted infrared signal sequence.

[0064] Specifically, the infrared signal evolution sequence is traversed in time order according to a unified time reference sequence, and the heading control degree of freedom corresponding to each acquisition time is read one by one; when the mapping result of the heading control degree of freedom at the corresponding acquisition time is that heading adjustment is allowed, the multi-channel infrared signal intensity at the corresponding acquisition time is retained; when the mapping result of the heading control degree of freedom at the corresponding acquisition time is that heading adjustment is restricted, it is not included in the subsequent resampling calculation, and the direction-restricted time index sequence is obtained.

[0065] Based on the directionally restricted time index sequence, the infrared signal evolution sequence is reconstructed into a time-increasing set of multi-channel infrared signal intensities according to the chronological order of the acquisition times in the directionally restricted time index sequence. After completing the directionally restricted resampling process, the multi-channel infrared signal intensities corresponding to each acquisition time in the infrared signal evolution sequence are arranged according to the channel identifier. The multi-channel infrared signal intensities corresponding to the directionally restricted time index sequence are then combined sequentially in chronological order to generate a restricted infrared signal sequence.

[0066] S3.2: Construct a spatial adjacency constraint matrix based on the constrained infrared signal sequence, and use the spatiotemporal-frequency joint embedding method to perform a unified representation of the co-evolution, generating a spatiotemporal-frequency coherence tensor.

[0067] Specifically, based on the restricted infrared signal sequence, the restricted infrared signal sequence is reconstructed into a two-dimensional arrangement structure with the acquisition time as the first dimension and the channel identifier as the second dimension according to the acquisition time sequence. A spatial adjacency constraint matrix is ​​constructed (representing whether there is a spatial adjacency relationship and the adjacency strength between two corresponding channels). In the time dimension of the spatial adjacency constraint matrix, the infrared signal intensity corresponding to each channel identifier is divided into multiple time windows, and the number of sign changes and amplitude changes of the infrared signal intensity are counted in each time window as frequency dimension representations.

[0068] The acquisition time, channel identifier, and frequency dimension representation of the corresponding time window are combined one-to-one and stacked in three dimensions according to the acquisition time order to construct the spatiotemporal frequency coherence tensor (including time dimension, channel dimension and frequency dimension).

[0069] It should be noted that the spatiotemporal-frequency joint embedding method is a multidimensional representation method that unifies and reconstructs information in the time dimension, channel dimension, and frequency dimension. By arranging the signal intensity changes in time sequence, maintaining the spatial distribution structure in the channel identification dimension, and extracting frequency features that reflect the rhythm of change within the time window, and combining them one-to-one and stacking them in three dimensions, a unified data carrier is provided for subsequent coherence analysis and interference component identification.

[0070] S3.3: Perform geometric propagation on the spatiotemporal frequency coherence tensor and execute propagation consistency analysis to obtain non-target interference components.

[0071] Specifically, the spatiotemporal frequency coordinates in the spatiotemporal frequency coherence tensor are extracted, and the corresponding tensor values ​​are used as the propagation start nodes. The process is advanced point by point in the time dimension according to the order of the acquisition time, and point by point is traversed in the channel and frequency dimensions. The direction of change of tensor values ​​at adjacent positions is compared, and the adjacency relationship with consistent direction of change is recorded.

[0072] Based on the adjacency relationship that maintains a consistent direction of change, the tensor values ​​corresponding to acquisition times T1, T2, T3, and T4 are considered continuous in the time dimension and expand in the channel and frequency dimensions. For example, in the spatiotemporal frequency coherence tensor corresponding to acquisition time T5, there are 4 channel components (channel 1, channel 2, channel 3, and channel 4) and 3 frequency components (frequency component 1, frequency component 2, and frequency component 3), and all channels and frequency positions have corresponding values. Therefore, the tensor values ​​are considered to expand in the channel and frequency dimensions. Positions that are discontinuous in the time dimension and do not expand in the channel and frequency dimensions are marked as propagation boundaries to obtain the continuously expanding propagation region.

[0073] The propagation region that cannot form a continuous extension in the channel dimension and frequency dimension will be extracted as a non-target interference component.

[0074] S3.4: Separate the non-target interference components in the spatiotemporal frequency coherence tensor and perform time continuity constraint integration to obtain the continuous signal change segment.

[0075] Specifically, the spatiotemporal frequency coherence tensor is marked according to the spatiotemporal frequency coordinate positions corresponding to the non-target interference components, and the tensor values ​​at the corresponding coordinate positions are removed from the spatiotemporal frequency coherence tensor, retaining the tensor values ​​that are not marked as non-target interference components, thus forming the target-related spatiotemporal frequency tensor region.

[0076] The target-related spatiotemporal frequency tensor region is traversed step-by-step according to the time dimension. The set of tensor values ​​that still maintain expansion in the channel and frequency dimensions at each acquisition time is counted, and the valid state markers of the corresponding acquisition time are recorded.

[0077] The valid status markers are arranged in the order of acquisition time. When there are valid status markers in adjacent acquisition times, they are continuously spliced ​​in the time dimension. When the valid status markers are interrupted between adjacent acquisition times, the interruption boundary is recorded. The valid status interval formed by continuous splicing in the time dimension is defined by the boundary between the start acquisition time and the end acquisition time, and then arranged and integrated in the time order to obtain the continuous signal change segment.

[0078] like Figure 6 As shown, the evolution of the coupled signal after logarithmic domain mapping and the lateral and heading offset components obtained by decomposition using the offset decoupling equation is presented. It can be observed that the coupling characteristics without decoupling exhibit an oscillating decay trend, with cross-influences between the signals in different dimensions, making it difficult to directly use for control quantity generation. By reducing the amplitude difference through logarithmic domain mapping and decomposing the signal structure using the offset decoupling equation, the coupled offset characteristics can be transformed into physically meaningful lateral and heading offset components. Both component curves show a monotonically converging trend over time, with the oscillation amplitude gradually decreasing. The results demonstrate that this invention, through logarithmic domain mapping and structured offset decoupling processing, achieves accurate mapping of signal structural features to control quantities, improving the stability and convergence speed of the robot's autonomous alignment process and enhancing control accuracy and robustness in complex environments.

[0079] S4: Based on the continuous signal variation segment, the infrared signal intensity is mapped to the logarithmic domain, and the offset decoupling equation is constructed to obtain the offset direction and offset degree.

[0080] S4.1: Perform logarithmic domain mapping processing on the intensity of multi-channel infrared signals within the continuous signal variation segment to generate a logarithmic domain infrared signal segment sequence.

[0081] Specifically, according to the range of acquisition times included in the continuous signal change segment, each acquisition time within the continuous signal change segment is traversed in time sequence, and the intensity of the corresponding multi-channel infrared signal in the infrared signal evolution sequence is read at each acquisition time position.

[0082] For each channel identifier corresponding to each acquisition time, the infrared signal intensity value is checked one by one to see if the infrared signal intensity value is positive. When the infrared signal intensity value is positive, the infrared signal intensity value is input into the natural logarithm function to obtain the infrared signal intensity logarithm value. When the infrared signal intensity value is zero or negative, the infrared signal intensity value of the corresponding channel identifier at the corresponding acquisition time is discarded.

[0083] The logarithmic values ​​of infrared signal intensity are arranged in the order of channel identifiers within each acquisition time, and then arranged in the time dimension in ascending order of acquisition time. The logarithmic form of multi-channel infrared signal intensity corresponding to all acquisition times is continuously combined to generate a logarithmic domain infrared signal segment sequence.

[0084] S4.2: Calculate the logarithmic difference change of the intensity of the multi-channel infrared signal in the logarithmic domain infrared signal segment sequence, and construct the multi-channel intensity gradient vector.

[0085] Specifically, the data is traversed in ascending order of the acquisition times recorded in the logarithmic domain infrared signal segment sequence, and the logarithmic values ​​of the multi-channel infrared signal intensity corresponding to two adjacent acquisition times are matched channel by channel. According to the channel identifier, the corresponding logarithmic value of the infrared signal intensity is read at the corresponding acquisition time position, and the logarithmic value of the infrared signal intensity corresponding to the same channel identifier is read at the previous acquisition time position.

[0086] The difference between the logarithmic value of the infrared signal intensity at the current acquisition time and the logarithmic value of the infrared signal intensity at the previous acquisition time is used as the logarithmic difference change at the corresponding channel identifier at the current acquisition time. Within each acquisition time, the corresponding logarithmic difference changes are combined according to the channel identifier to form a multi-channel intensity gradient vector.

[0087] S4.3: Based on the multi-channel intensity gradient vector, the offset decoupling equation is constructed by odd-even grouping and linear combination according to the channel identifier, and decomposed into lateral offset component and heading offset component.

[0088] Specifically, read the multi-channel intensity gradient vector components of all channel identifiers corresponding to the current acquisition time, and number them in ascending order according to the arrangement number of the channel identifiers; count the multi-channel intensity gradient vector components corresponding to the channel identifiers with odd numbers to obtain the first set of accumulated values; count the multi-channel intensity gradient vector components corresponding to the channel identifiers with even numbers to obtain the second set of accumulated values.

[0089] The difference between the first and second accumulated values ​​is taken as the lateral offset component; the sum of the first and second accumulated values ​​is taken as the heading offset component; the lateral offset component and heading offset component are statistically analyzed at each acquisition time within the continuous signal change segment to generate the offset decoupling equation and obtain the lateral offset component and heading offset component.

[0090] It should be noted that the offset decoupling equation is a linear combination expression based on multi-channel intensity gradient vectors. By grouping the multi-channel intensity gradient vectors according to the channel identifier, different spatial offset effects in the same observation are separated, thereby obtaining the lateral offset component and the heading offset component respectively. This achieves the independent representation between different offset dimensions and provides a structured expression basis for subsequent offset direction determination and offset degree quantification.

[0091] S4.4: The lateral offset component and the heading offset component are integrated over time and their signs are determined within the continuous signal variation segment to obtain the offset direction and offset degree.

[0092] Specifically, according to the range of acquisition times included in the continuous signal change segment, the lateral offset components corresponding to each acquisition time are read one by one in chronological order; the sum of the lateral offset components corresponding to all acquisition times within the continuous signal change segment is calculated to obtain the lateral cumulative offset; the lateral offset direction is determined according to the sign of the lateral cumulative offset, when the lateral cumulative offset is positive, it is determined to be a positive lateral offset, when the lateral cumulative offset is negative, it is determined to be a negative lateral offset, and the absolute value of the lateral cumulative offset is used as the degree of lateral offset.

[0093] Within the same continuous signal variation segment, the sum of the heading offset components corresponding to each acquisition moment in chronological order is calculated to obtain the cumulative heading offset; the heading offset direction is determined according to the sign of the cumulative heading offset, and the absolute value of the cumulative heading offset is taken as the degree of heading offset.

[0094] By combining the lateral offset direction and the lateral offset degree, and the heading offset direction and the heading offset degree, the offset direction and the offset degree corresponding to the continuous signal change segment are obtained.

[0095] S5: Map the offset direction and offset degree to the heading correction and position correction, and perform closed-loop correction to generate alignment control commands.

[0096] S5.1: The direction and magnitude coupling mapping algorithm is used to generate heading and position corrections by using the offset direction as the sign constraint and the offset degree as the magnitude scale.

[0097] Specifically, at the end of the continuous signal change segment, the lateral offset direction, lateral offset degree, and heading offset direction and degree are read; the lateral offset direction and heading offset direction are symbolized, and when the lateral offset direction is positive, it is identified as a positive symbol mark, and when the lateral offset direction is negative, it is identified as a negative symbol mark; when the heading offset direction is positive, it is identified as a positive symbol mark, and when the heading offset direction is negative, it is identified as a negative symbol mark.

[0098] The degree of lateral offset is used as the source of position correction amplitude, and the degree of heading offset is used as the source of heading correction amplitude. The expression of position correction amplitude and heading correction amplitude are obtained. The expression of position correction amplitude is symbolically bound to the symbol corresponding to the direction of lateral offset to form the position correction amount. The expression of heading correction amplitude is symbolically bound to the symbol corresponding to the direction of heading offset to form the heading correction amount.

[0099] It should be noted that the direction-amplitude coupling mapping algorithm is a mapping method that unifies the expression of offset direction and offset degree. By embedding the offset direction as a symbolic constraint into the positive and negative attributes of the output quantity, and embedding the offset degree as an amplitude scale into the numerical value of the output quantity, the heading correction and position correction quantities carry both direction and amplitude information during the generation process, thus realizing the synchronous coupling expression of direction determination and amplitude adjustment.

[0100] S5.2: Based on the heading correction and position correction, a stable correction pair is generated by performing coupled constraint processing through an interlocking vibration suppression adjustment algorithm.

[0101] Specifically, the heading correction and position correction values ​​are arranged sequentially according to the acquisition time in the continuous signal change segment, forming a heading correction value sequence and a position correction value sequence. For each acquisition time, the heading correction value is extracted from the heading correction value sequence and the position correction value is extracted from the position correction value sequence, and the sign is determined respectively. When the heading correction value and the position correction value are positive, the output is positive; when the heading correction value and the position correction value are negative, the output is negative; when the heading correction value and the position correction value are zero, the output is zero.

[0102] Interlock consistency processing is performed on the heading correction sign determination and position correction sign determination corresponding to the acquisition time. When the heading correction sign determination output and the position correction sign determination have opposite direction combinations, the heading correction value and the position correction value are set to zero at the same time; when the heading correction sign determination output and the position correction sign determination have the same direction, the heading correction value and the position correction value remain unchanged.

[0103] Take the absolute values ​​of the heading correction and the position correction respectively, and take the minimum value as the uniform amplitude; combine the heading correction sign determination and the position correction sign determination corresponding to the acquisition time into heading stability correction and position stability correction, and arrange them in pairs in time order to generate stability correction pairs.

[0104] It should be noted that the interlocked vibration suppression adjustment algorithm updates the heading correction and position correction in a time sequence by mutually restricting each other. When the direction of change of the heading correction and the direction of change of the position correction show an opposite superposition trend, the change amplitude is progressively reduced according to the corresponding offset direction and offset degree, so that the change of correction between consecutive acquisition times keeps the transition in the same direction, avoiding the oscillation amplification phenomenon of the heading correction and position correction during the closed-loop correction process, and achieving stable convergent correction output.

[0105] S5.3: Recursively update the stable correction pairs and perform boundary limiting processing to generate alignment control commands.

[0106] Specifically, the stable correction pairs are read sequentially according to the acquisition time sequence. At each acquisition time position, the current stable correction pair is sequentially superimposed with the heading and position corrections in the alignment control command corresponding to the previous acquisition time. For example, if the current stable correction pair has a heading correction of 2 and a position correction of 3, and the alignment control command corresponding to the previous acquisition time has a heading correction of 5 and a position correction of 4, then the superposition results in a heading correction of 7 and a position correction of 7, forming the progressive update correction for the current acquisition time.

[0107] The heading and position corrections corresponding to all acquisition times within the continuous signal variation segment are traversed and statistically analyzed, and the maximum and minimum values ​​are recorded respectively to form the allowable output ranges for heading and position corrections. The progressive update correction at the current acquisition time is compared with the upper and lower boundaries. When the progressive update correction exceeds the allowable output range, the progressive update correction is truncated to the corresponding boundary value. When the progressive update correction is within the allowable output range, the value remains unchanged. The heading and position corrections are arranged in chronological order and combined to generate alignment control commands.

[0108] In summary, this invention achieves collaborative evolution analysis of multi-channel infrared signals in the time, space, and frequency dimensions by constructing a spatiotemporal frequency coherence tensor and using a geometric propagation algorithm to separate non-target interference components, thereby improving the continuity and reliability of industrial control software in complex environments. Furthermore, by using logarithmic domain mapping and offset decoupling equations, the coupled signal is decomposed into lateral and heading offset components, achieving precise conversion of signal structural features into control quantities and improving the stability and convergence of the robot's autonomous alignment process.

[0109] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for processing infrared alignment signal data for robots, characterized in that, include: The infrared signal intensity of the infrared alignment signals from multiple channels is collected and time-aligned to obtain the infrared signal evolution sequence. Based on the infrared signal evolution sequence, the changing trend of infrared signal intensity is analyzed, the heading adjustment state is determined, and the heading control degrees of freedom are obtained. Based on the heading control degrees of freedom, a spatiotemporal frequency coherence tensor of the infrared alignment signal is constructed, and a geometric propagation algorithm is used to identify and separate non-target interference components to obtain continuous signal variation segments. Based on the continuous signal variation segment, the infrared signal intensity is mapped to the logarithmic domain, and an offset decoupling equation is constructed to obtain the offset direction and offset degree; The offset direction and offset degree are mapped to heading correction and position correction, and closed-loop correction is performed to generate alignment control commands.

2. The infrared alignment signal data processing method for robots as described in claim 1, characterized in that: The infrared signal recording set includes channel identifier, acquisition time, and infrared signal strength; The infrared signal intensity of the infrared alignment signals from multiple channels is collected, and time alignment is performed to obtain the infrared signal evolution sequence. The specific steps are as follows: The infrared signal intensity of infrared alignment signals from multiple channels is collected, and the corresponding collection time is recorded to obtain an infrared signal record set. Align the channel identifier with the acquisition time and map the infrared signal intensity to the same time base to generate a multi-channel infrared signal sequence.

3. The infrared alignment signal data processing method for robots as described in claim 2, characterized in that: The infrared signal evolution sequence is obtained by arranging the infrared signal intensity in ascending order according to the time sequence of the multi-channel infrared signal sequence.

4. The infrared alignment signal data processing method for robots as described in claim 1, characterized in that, The specific steps for analyzing the changing trend of infrared signal intensity based on the infrared signal evolution sequence are as follows: Based on the infrared signal evolution sequence, abrupt segments are marked and removed using the sliding window decomposition method to obtain a set of continuous and effective signal segments; Calculate the multi-channel signal consistency score within a set of continuous valid signal segments, perform spatiotemporal frequency coherence assessment, and obtain coherence labels; The same coherence markers are merged into coherent segments and arranged in chronological order to generate a coherent segment sequence.

5. The infrared alignment signal data processing method for robots as described in claim 4, characterized in that: The heading control degrees of freedom are obtained by comparing the persistence and consistency of the coherent segment sequence, generating a heading adjustment state segment sequence, and performing adjustment state mapping.

6. The infrared alignment signal data processing method for robots as described in claim 1, characterized in that, The specific steps for constructing the spatiotemporal frequency coherence tensor of the infrared alignment signal based on the heading control degrees of freedom are as follows: The infrared signal evolution sequence is resampled and rearranged in a directionally restricted manner according to the heading control degree of freedom to generate a restricted infrared signal sequence. A spatial adjacency constraint matrix is ​​constructed based on a constrained infrared signal sequence. A unified representation of the co-evolution is generated by using a spatiotemporal-frequency joint embedding method.

7. The infrared alignment signal data processing method for robots as described in claim 6, characterized in that, The specific steps for obtaining the continuously changing signal segment are as follows: Geometric propagation is performed on the spatiotemporal frequency coherence tensor, and propagation consistency analysis is conducted to obtain non-target interference components; Non-target interference components in the spatiotemporal frequency coherence tensor are separated and integrated with temporal continuity constraints to obtain continuous signal variation segments.

8. The infrared alignment signal data processing method for robots as described in claim 7, characterized in that, The process of mapping infrared signal intensity to the logarithmic domain based on continuous signal variation segments and constructing an offset decoupling equation to obtain the offset direction and degree is as follows: Logarithmic domain mapping is performed on the intensity of multi-channel infrared signals within a continuous signal variation segment to generate a logarithmic domain infrared signal segment sequence. Calculate the logarithmic difference of the intensity of multi-channel infrared signals in the sequence of logarithmic domain infrared signals, and construct the multi-channel intensity gradient vector; Based on the multi-channel intensity gradient vector, the offset decoupling equation is constructed by odd-even grouping and linear combination according to the channel identifier, and decomposed into lateral offset component and heading offset component. The lateral offset component and the heading offset component are integrated over time and their signs are determined within the continuous signal variation segment to obtain the offset direction and offset degree.

9. The infrared alignment signal data processing method for robots as described in claim 8, characterized in that: The process of mapping the offset direction and offset degree to heading correction and position correction refers to using a direction amplitude coupling mapping algorithm to generate heading correction and position correction by using the offset direction as a symbol constraint and the offset degree as an amplitude scale.

10. The infrared alignment signal data processing method for robots as described in claim 9, characterized in that, The specific steps for generating alignment control commands are as follows: Based on the heading correction and position correction, a stable correction pair is generated by performing coupled constraint processing through an interlocking vibration suppression adjustment algorithm. The stable correction pairs are recursively updated and boundary limiting is performed to generate alignment control commands.

Images