Training analysis method and system for animal cognitive movement

Abstract

The invention discloses a training analysis method and system for animal cognitive movement, and particularly relates to the technical field of training analysis, and the method comprises the steps: collecting original movement behavior data generated by an animal in a cognitive task conversion process, and carrying out the feature segmentation to generate a plurality of movement behavior subsequences; in combination with time nodes of cognitive task type changes, task labels of the motion behavior subsequences are extracted, behavior characteristic differences between adjacent cognitive task subsequences are analyzed, and a cross-task difference characteristic matrix is constructed; establishing a cognitive migration interference identification mechanism based on a reverse cognitive migration interference index, and identifying an abnormal behavior fragment and a key task conversion node; constructing a multi-dimensional interference evaluation model in combination with historical cognitive training data, and evaluating the cognitive migration interference degree of the animal at the current training stage; according to the method, the training task sequence is adjusted according to the matching result of the cognitive migration interference degree and the preset interference threshold value, and real-time adaptation and interference suppression of the animal cognitive migration state are achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of training analysis, and more specifically, to a training analysis method and system for animal cognitive movement. Background Art

[0002] Animal cognitive movement training has important application values in the fields of neuroscience, ethology and medical research. However, in the actual training process, when an animal continuously performs different types of cognitive tasks, the cognitive patterns or strategies formed in its early stage often cannot be adjusted in a timely and effective manner, resulting in negative cognitive transfer, that is, the animal continuously uses the old cognitive strategies and cannot smoothly adapt to the new task requirements. This reverse cognitive transfer interference phenomenon leads to a decrease in the stability and accuracy of animal behavior data, significantly affecting the training effect and the reliability of experimental data. Existing training analysis methods lack an effective identification and coping mechanism for this interference effect.

[0003] To solve the above problems, a training analysis method and system for animal cognitive movement are provided herein. Summary of the Invention

[0004] To overcome the above defects of the prior art, embodiments of the present invention provide a training analysis method and system for animal cognitive movement to solve the problems raised in the above background art.

[0005] To achieve the above object, the present invention provides the following technical solutions: A training analysis method for animal cognitive movement, comprising the following steps: Obtain the original movement behavior data generated by an animal during the process of cognitive task transition, perform feature segmentation processing on the original movement behavior data to generate a plurality of movement behavior subsequences; According to the time nodes when the animal accepts the changes in the types of cognitive tasks, extract the cognitive task labels corresponding to each movement behavior subsequence, and analyze the behavioral feature differences between adjacent cognitive task subsequences to generate a cross-task difference feature matrix; Establish a reverse cognitive transfer interference index, identify abnormal movement behavior segments deviating from the normal cognitive pattern according to the cross-task difference feature matrix, and determine the key task conversion nodes that generate reverse cognitive transfer interference; Construct a multi-dimensional evaluation model for reverse cognitive transfer interference based on historical cognitive training data, and evaluate the degree of cognitive transfer interference in the current training stage of the animal according to the key task conversion nodes and the corresponding abnormal movement behavior segments; Match the degree of cognitive transfer interference in the current training stage of the animal with a preset interference threshold, and adjust the sequence arrangement of cognitive tasks according to the matching result.

[0006] In a preferred embodiment, the original motion behavior data generated by the animal during the cognitive task transition is obtained, and the original motion behavior data is subjected to feature segmentation processing to generate a plurality of motion behavior subsequences. Specifically: Multi-channel parallel acquisition of the original motion behavior data of the animal in different cognitive task execution stages; The collected original motion behavior data is segmented according to the cognitive task trigger time node and divided into a plurality of initial behavior paragraph data sets; Based on the initial behavior paragraph data set, the behavior structure feature indicators in the initial behavior paragraph data are extracted to form a structured behavior feature representation; According to the change pattern of the behavior structure feature indicators in the structured behavior feature representation, a sliding window method is used for fine-grained segmentation to generate a set of motion behavior subsequences that meet the consistency of a single behavior feature.

[0007] In a preferred embodiment, according to the time node when the animal accepts the change of the cognitive task type, the cognitive task label corresponding to each motion behavior subsequence is extracted, and the behavior feature differences between adjacent cognitive task subsequences are analyzed to generate a cross-task difference feature matrix. Specifically: The time range corresponding to each motion behavior subsequence is matched with the time node of the cognitive task type change to determine the cognitive task type label corresponding to each motion behavior subsequence; According to the cognitive task type label of each motion behavior subsequence, two motion behavior subsequences arranged continuously in the time dimension are defined as a pair of adjacent cognitive task subsequences; For each pair of adjacent cognitive task subsequences, the behavior feature difference index corresponding to each pair of adjacent cognitive task subsequences is calculated; Based on each pair of adjacent cognitive task subsequences and the corresponding behavior feature difference index, a cross-task difference feature matrix is constructed.

[0008] In a preferred embodiment, a reverse cognitive transfer interference index is established, and abnormal motion behavior segments deviating from the normal cognitive mode are identified according to the cross-task difference feature matrix, and the key task conversion nodes that generate reverse cognitive transfer interference are determined. Specifically: Based on the cross-task difference feature matrix, a calculation rule for the reverse cognitive transfer interference index is established; According to the calculation rule of the reverse cognitive transfer interference index, the behavior feature difference indexes in the cross-task difference feature matrix are weighted and calculated, and the reverse cognitive transfer interference index value is output; According to the magnitude relationship between the reverse cognitive transfer interference index value and the preset cognitive mode deviation threshold, it is determined whether the motion behavior subsequence belongs to an abnormal motion behavior segment deviating from the normal cognitive mode; Determine the key task transition nodes that cause reverse cognitive transfer interference according to the adjacent cognitive task subsequences corresponding to abnormal movement behavior segments.

[0009] In a preferred embodiment, a multi-dimensional evaluation model of reverse cognitive transfer interference is constructed based on historical cognitive training data. According to the key task transition nodes and the corresponding abnormal movement behavior segments, the degree of cognitive transfer interference in the current training stage of the animal is evaluated. Specifically: Extract the cognitive task type labels corresponding to the key task transition nodes and the behavioral feature difference indicators corresponding to the abnormal movement behavior segments during the historical cognitive training process; Generate a multi-dimensional evaluation model of reverse cognitive transfer interference based on the influence weights of the cognitive task type labels and the behavioral feature difference indicators on reverse cognitive transfer interference; Input the cognitive task type labels and behavioral feature difference indicators in the current training stage into the multi-dimensional evaluation model of reverse cognitive transfer interference, and output the cognitive transfer interference value of the animal in the current training stage.

[0010] In a preferred embodiment, the degree of cognitive transfer interference in the current training stage of the animal is matched with a preset interference threshold, and the sequence arrangement of the cognitive tasks is adjusted according to the matching result. Specifically: Obtain the preset interference threshold, and compare the cognitive transfer interference value of the animal in the current training stage with the preset interference threshold; When the cognitive transfer interference value of the animal in the current training stage is greater than the preset interference threshold, reorder the cognitive training task sequence according to the similarity of the cognitive task type labels from high to low, and insert a transition training link between adjacent cognitive training tasks; When the cognitive transfer interference value of the animal in the current training stage is less than or equal to the preset interference threshold, keep the cognitive training task sequence unchanged.

[0011] On the other hand, the present invention provides a training analysis system for animal cognitive movement, including a behavior acquisition module, a label extraction module, an interference recognition module, an interference evaluation module, and a sequence adjustment module; The behavior acquisition module acquires the original movement behavior data generated by the animal during the cognitive task transition, performs feature segmentation processing on the original movement behavior data, and generates multiple movement behavior subsequences; The label extraction module extracts the cognitive task labels corresponding to each movement behavior subsequence according to the time nodes when the animal receives changes in the cognitive task types, and analyzes the behavioral feature differences between adjacent cognitive task subsequences to generate a cross-task difference feature matrix; The interference recognition module establishes an index of reverse cognitive transfer interference, identifies abnormal movement behavior segments that deviate from the normal cognitive mode according to the cross-task difference feature matrix, and determines the key task transition nodes that cause reverse cognitive transfer interference; The interference evaluation module constructs a multi-dimensional evaluation model for reverse cognitive transfer interference based on historical cognitive training data, and evaluates the degree of cognitive transfer interference in the current training stage of the animal according to the key task conversion nodes and the corresponding abnormal movement behavior segments. The sequence adjustment module matches the degree of cognitive transfer interference in the current training stage of the animal with a preset interference threshold, and adjusts the sequence arrangement of cognitive tasks according to the matching result.

[0012] The technical effects and advantages of the training analysis method and system for animal cognitive movement of the present invention: By constructing the correspondence between the movement behavior subsequences and the cognitive task type labels, and combining the behavioral structure feature difference indicators, the abnormal behavior performance of the animal in the cognitive mode switching can be accurately identified; by constructing the cross-task difference feature matrix and the reverse cognitive transfer interference index, the key task conversion nodes can be accurately located, and a multi-dimensional evaluation model is established based on historical cognitive training data to quantify the interference degree of the animal in the current stage. The real-time determination of cognitive transfer interference is realized, and the sorting of subsequent training tasks is adjusted according to the matching result with the cognitive threshold, improving the training continuity and adaptability, and effectively improving the efficiency and stability of animal cognitive training. Brief Description of the Drawings

[0013] Figure 1 It is a schematic diagram of the training analysis method for animal cognitive movement of the present invention; Figure 2 It is a schematic diagram of the structure of the training analysis system for animal cognitive movement of the present invention. Detailed Embodiments

[0014] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0015] Embodiment 1 Figure 1 The training analysis method for animal cognitive movement of the present invention is given, which includes the following steps: Obtain the original movement behavior data generated by the animal during the cognitive task transition, and perform feature segmentation processing on the original movement behavior data to generate multiple movement behavior subsequences; According to the time nodes when the animal receives the change of cognitive task types, extract the cognitive task labels corresponding to each movement behavior subsequence, and analyze the behavioral feature differences between adjacent cognitive task subsequences to generate a cross-task difference feature matrix; Establish a reverse cognitive transfer interference index, identify abnormal segments of motor behavior that deviate from the normal cognitive pattern according to the cross-task difference feature matrix, and determine the key task transition nodes that generate reverse cognitive transfer interference; Construct a multi-dimensional evaluation model of reverse cognitive transfer interference based on historical cognitive training data, and evaluate the degree of cognitive transfer interference in the current training stage of the animal according to the key task transition nodes and the corresponding abnormal segments of motor behavior; Match the degree of cognitive transfer interference in the current training stage of the animal with a preset interference threshold, and adjust the sequence arrangement of cognitive tasks according to the matching result.

[0016] Specifically, obtain the original motor behavior data generated by the animal during the cognitive task transition, perform feature segmentation processing on the original motor behavior data, and generate multiple sub-sequences of motor behavior, including: Perform multi-channel parallel acquisition on the original motor behavior data of the animal in different cognitive task execution stages; Specifically, build a trajectory recording device, a speed sensing device and a time recording device on the animal cognitive motor training platform. The trajectory recording device is used to capture in real time the change sequence of two-dimensional or three-dimensional position coordinate points generated by the animal in the cognitive task training scenario; the speed sensing device is used to obtain the real-time movement speed value of the animal at each movement sampling moment; the time recording device is used to record the time point when each group of sampling data occurs, for time dimension segmentation.

[0017] All acquisition devices work together through a synchronous control unit to ensure that within the same sampling period, the trajectory, speed and time information are in one-to-one correspondence, generating a unified structured original motor behavior data stream. The original motor behavior data stream provides basic support for task recognition and behavior segmentation.

[0018] Segment the collected original motor behavior data according to the cognitive task trigger time node, and divide it into multiple initial behavior paragraph data sets; Specifically, based on the task trigger time node in the cognitive training task setting, sequentially segment the original motor behavior data according to the time axis. The start and end times of each segmented segment correspond to the start and end times of a cognitive task respectively, ensuring that each segment of data contains the complete behavior trajectory of the animal in a single cognitive task execution stage.

[0019] Exemplarily, the experimental setting includes three consecutive cognitive tasks, namely a path recognition task, a target discrimination task and a spatial jump task, and the task trigger times are time point one, time point two and time point three in sequence. Then the original motor behavior data is divided into three initial behavior paragraph data according to the task trigger time, corresponding to all the motor behavior data during the execution of the three cognitive tasks.

[0020] In each initial behavior paragraph data, the trajectory coordinate sequence, speed sequence, and corresponding time sequence within the corresponding time interval are retained simultaneously, forming a unified data triple representation structure for structural feature extraction.

[0021] Based on the initial behavior paragraph data set, extract the behavioral structure feature indicators in the initial behavior paragraph data to form a structured behavioral feature representation; Specifically, perform structured processing on each initial behavior paragraph data and extract four types of behavioral structure feature indicators: Degree of trajectory variation: By calculating the average distance between all position points in the initial behavior paragraph data and the center line of the fitted trajectory, the tortuosity of the animal's movement trajectory is measured. The more tortuous the animal's movement trajectory, the greater the average distance, indicating that there are more directional changes during the movement process.

[0022] Speed fluctuation characteristics: Calculate the standard deviation of the speed data sequence in the initial behavior paragraph data to reflect the stability of the animal's movement speed during the cognitive task stage. The larger the standard deviation, the more frequent the acceleration or deceleration behaviors of the animal during task execution.

[0023] Behavior duration: Calculated from the difference between the start time point and the end time point of the initial behavior paragraph data, representing the time length used by the animal to complete the cognitive task.

[0024] Displacement amplitude: By calculating the straight-line distance between the start point and the end point in the initial behavior paragraph data, it is used to represent the overall movement space range spanned by the animal during the completion of the current cognitive task.

[0025] Summarize the above four types of behavioral structure feature indicators to form a structured behavioral feature representation of the initial behavior paragraph data. Each initial behavior paragraph data has a set of structured behavioral features.

[0026] According to the change pattern of the behavioral structure feature indicators in the structured behavioral feature representation, use a sliding window method for fine-grained segmentation to generate a set of movement behavior subsequences that meet the consistency of a single behavioral feature; Specifically, after completing the construction of the structured behavioral feature representation, use a sliding window method to perform fine-grained analysis on each initial behavior paragraph data. Each sliding window covers a set of continuous trajectory point sequences, speed data sequences, and time sequences in the initial behavior paragraph data, and extracts the above four types of behavioral structure feature indicators in units of sliding windows.

[0027] By analyzing the change patterns of feature indicators between consecutive sliding windows, it is determined whether the behavioral characteristics are stable. When the trajectory variation degree, speed fluctuation characteristics, behavioral duration, and displacement amplitude change between two adjacent sliding windows are all within the set feature consistency threshold, it is determined that the initial behavioral paragraph data has consistent behavioral characteristics.

[0028] In the entire initial behavioral paragraph data, all sliding window regions that meet the feature consistency are merged to form multiple motion behavior subsequences that meet the single behavioral feature consistency. Each motion behavior subsequence has continuous and stable trajectory forms and speed characteristics within the corresponding time range, and can be used as the basic unit for cognitive task label matching and interference index calculation.

[0029] Specifically, according to the time nodes of the changes in the types of cognitive tasks received by the animal, the cognitive task labels corresponding to each motion behavior subsequence are extracted, and the behavioral feature differences between adjacent cognitive task subsequences are analyzed to generate a cross-task difference feature matrix, including: Match the time range corresponding to each motion behavior subsequence with the time nodes of the changes in the types of cognitive tasks to determine the cognitive task type label corresponding to each motion behavior subsequence; Specifically, based on the cognitive task execution plan configured in the experimental setting stage, the actual trigger time nodes of each cognitive task are known, forming a cognitive task time node sequence. Each motion behavior subsequence in the set of motion behavior subsequences has a clear start time and end time, which are determined by the collected time information.

[0030] Match the start and end time ranges of the motion behavior subsequence with the cognitive task time node sequence one by one to determine whether the time interval of each motion behavior subsequence completely falls within the start and end time ranges of the cognitive task. If the time interval of each motion behavior subsequence completely falls within the start and end time ranges of the cognitive task, the corresponding relationship between the motion behavior subsequence and the cognitive task can be determined.

[0031] For example, in the cognitive task sequence configured in the training process, the tasks are path recognition task, target discrimination task, and space jump task, and the task trigger times are the first time point, the second time point, and the third time point respectively. If the start time of the motion behavior subsequence is greater than or equal to the first time point and the end time is less than the second time point, it is determined that the motion behavior subsequence corresponds to the path recognition task.

[0032] Through the above method, the binding between the motion behavior subsequence and the cognitive task type label is completed, and a unique cognitive task type label is attached to each motion behavior subsequence to form a set of labeled motion behavior subsequences.

[0033] According to the cognitive task type labels of each motion behavior subsequence, two motion behavior subsequences arranged continuously in the time dimension are defined as a pair of adjacent cognitive task subsequences; Specifically, in the set of labeled motion behavior subsequences, each motion behavior subsequence is arranged in chronological order, and every two continuously occurring motion behavior subsequences with different labels are defined as a pair of adjacent cognitive task subsequences.

[0034] For example, if the labels of the subsequences arranged in chronological order are path recognition task, target discrimination task, and spatial jump task in sequence, then the path recognition task and the target discrimination task form the first pair of adjacent cognitive task subsequences, and the target discrimination task and the spatial jump task form the second pair of adjacent cognitive task subsequences.

[0035] Each pair of adjacent cognitive task subsequences constitutes a difference analysis unit for processing feature difference indicators.

[0036] For each pair of adjacent cognitive task subsequences, the corresponding behavior feature difference indicators of each pair of adjacent cognitive task subsequences are obtained through calculation; Specifically, for each pair of adjacent cognitive task subsequences, the behavior structure feature indicators are extracted in turn. The behavior structure feature indicators include trajectory variation degree, speed fluctuation feature, behavior duration, and displacement amplitude.

[0037] The values of the two motion behavior subsequences in each pair of adjacent cognitive task subsequences on the same type of behavior structure feature indicators are calculated for the difference to obtain the difference value of the behavior structure feature indicators, that is, the behavior feature difference indicator. For example, if the trajectory variation degree of the first motion behavior subsequence is the first value and the trajectory variation degree of the second motion behavior subsequence is the second value, then the difference in the trajectory variation degree of the adjacent cognitive task subsequences is the absolute value of the first value minus the second value. And so on, the calculation of the speed fluctuation feature difference, behavior duration difference, and displacement amplitude difference is completed in turn to obtain the set of behavior feature difference indicators corresponding to the adjacent cognitive task subsequences.

[0038] The above set of behavior feature difference indicators is used as the behavior difference representation of the adjacent cognitive task subsequences, providing basic variables for interference recognition and modeling.

[0039] Based on each pair of adjacent cognitive task subsequences and the corresponding behavior feature difference indicators, a cross-task difference feature matrix is constructed; Specifically, taking all pairs of adjacent cognitive task subsequences as rows and the types of behavioral feature difference indicators as columns, a cross-task difference feature matrix is constructed. Each row in the cross-task difference feature matrix corresponds to a pair of adjacent cognitive task subsequences, and each column corresponds in sequence to the difference in trajectory variation degree, the difference in speed fluctuation feature, the difference in behavior duration, and the difference in displacement amplitude. The value of each cell in the cross-task difference feature matrix is the difference value of the calculated behavioral structure feature indicators.

[0040] Specifically, a reverse cognitive transfer interference index is established. Based on the cross-task difference feature matrix, abnormal segments of movement behavior that deviate from the normal cognitive pattern are identified, and the key task transition nodes that generate reverse cognitive transfer interference are determined, including: Based on the cross-task difference feature matrix, a calculation rule for the reverse cognitive transfer interference index is established; Specifically, the behavioral feature difference indicators included in the cross-task difference feature matrix are defined standardly. For each type of behavioral structure feature difference indicator, a weight for the degree of behavior deviation is set. The weight for the degree of behavior deviation is determined according to the range of the historical behavioral difference stable interval statistically obtained in the pre-experiment stage. The higher the weight for the degree of behavior deviation, the stronger the sensitivity of the feature difference to cognitive transfer interference. For example, in most experimental animals, the change in the speed fluctuation feature has a greater impact on task cognitive switching than the change in the displacement amplitude, so the weight value corresponding to the speed fluctuation feature difference indicator is greater than the weight value corresponding to the displacement amplitude difference indicator.

[0041] The calculation rule for the reverse cognitive transfer interference index is: multiply the absolute value of each type of behavioral feature difference indicator by the corresponding weight for the degree of behavior deviation, and sum up all the product results. The obtained sum is the interference index value of the adjacent cognitive task subsequences in the dimension of behavioral structure features, which is used to measure the behavioral stability and consistency during the cognitive task switching process.

[0042] According to the calculation rule of the reverse cognitive transfer interference index, the behavioral feature difference indicators in the cross-task difference feature matrix are weighted and calculated to output the reverse cognitive transfer interference index value; Specifically, for each row in the cross-task difference feature matrix, that is, for each pair of adjacent cognitive task subsequences, the output process of the reverse cognitive transfer interference index value is performed according to the calculation rule of the reverse cognitive transfer interference index. The specific process includes: Read the absolute value of the difference in trajectory variation degree, the absolute value of the difference in speed fluctuation feature, the absolute value of the difference in behavior duration, and the absolute value of the difference in displacement amplitude.

[0043] The absolute values ​​of the differences in trajectory variation, speed fluctuation characteristics, behavior duration, and displacement amplitude are multiplied by their corresponding weight values, and the resulting products are recorded as four intermediate values.

[0044] The four intermediate values ​​are added together as the reverse cognitive transfer interference index value corresponding to the adjacent cognitive task subsequences.

[0045] According to the magnitude relationship between the reverse cognitive transfer interference index value and the preset cognitive mode deviation threshold, it is determined whether the motor behavior subsequence belongs to an abnormal motor behavior segment that deviates from the normal cognitive mode; Specifically, a cognitive mode deviation threshold is set to determine whether the behavioral deviation of the animal during cognitive task switching reaches the interference standard. The setting of the cognitive mode deviation threshold refers to the average behavioral fluctuation range of the experimental subjects, and the limit value that can represent the upper limit of normal cognitive behavior fluctuation is determined based on the statistical results of previous experiments.

[0046] The reverse cognitive transfer interference index value of each pair of adjacent cognitive task subsequences is compared with the cognitive mode deviation threshold. If the reverse cognitive transfer interference index value is greater than the cognitive mode deviation threshold, it is considered that the adjacent cognitive task subsequences have structural behavior drift during the task switching process, and there is a cognitive transfer disorder phenomenon, and the corresponding subsequent motor behavior subsequence is marked as a motor behavior abnormal segment.

[0047] Abnormal motor behavior episodes indicate that animals cannot effectively adapt to new task scenarios after cognitive switching.

[0048] According to the adjacent cognitive task subsequences corresponding to the abnormal motor behavior segments, the key task transition nodes that cause reverse cognitive transfer interference are determined; Specifically, each abnormal motor behavior segment originates from the transition stage from the previous cognitive task to the next cognitive task. The cognitive task switching time node in the transition stage constitutes the key task switching node. According to the task switching time point of the next subsequence corresponding to the abnormal motor behavior segment in the original cognitive task plan, the task switching boundary to which it belongs is determined, which is the key time node for the generation of reverse cognitive transfer interference.

[0049] All identified key task transition nodes will be summarized into a set of key time points for interference degree modeling, cognitive task sequence optimization and individual adaptability assessment.

[0050] Specifically, a multidimensional evaluation model of reverse cognitive transfer interference is constructed based on historical cognitive training data. According to the key task transition nodes and the corresponding abnormal movement behavior fragments, the degree of cognitive transfer interference in the current training stage of the animal is evaluated, including: Extract the cognitive task type labels corresponding to the key task transition nodes and the behavioral feature difference indicators corresponding to the abnormal motor behavior segments during the historical cognitive training process; Specifically, call the historical cognitive training data from the animal cognitive movement training platform. The historical cognitive training data includes the key task transition nodes and the corresponding abnormal motor behavior segments recorded in multiple training cycles.

[0051] For each key task transition node, extract the cognitive task type label at the time of its occurrence, that is, the specific type combination information of the previous cognitive task and the subsequent cognitive task represented by the key task transition node. For example, the path recognition task switches to the spatial jump task. Uniformly encode the cognitive task type labels to be used as one of the input dimensions for modeling.

[0052] For the abnormal motor behavior segments corresponding to each key task transition node, extract the behavioral feature difference indicators of the abnormal motor behavior segments.

[0053] Combine the cognitive task type labels and the behavioral feature difference indicators into complete feature entries, and summarize them to form a historical reverse cognitive transfer interference basic dataset for training a multi-dimensional evaluation model.

[0054] Generate a multi-dimensional evaluation model for reverse cognitive transfer interference based on the influence weights of the cognitive task type labels and the behavioral feature difference indicators on reverse cognitive transfer interference; Specifically, based on the cognitive task type labels and the behavioral feature difference indicators, adopt a weighted feature fusion strategy to construct a multi-dimensional evaluation model. The multi-dimensional evaluation uses the task switching type as a classification variable and the behavioral structure feature difference indicators as input variables to comprehensively evaluate the degree of reverse cognitive transfer interference.

[0055] Statistically calculate the mean and range of the behavioral structure feature difference indicators corresponding to each type of task switching type in all historical samples to analyze the contribution sensitivity of different task switching types to the interference risk.

[0056] According to the statistical results, respectively determine the specific weighting factors for the differences in trajectory variation degree, speed fluctuation characteristics, behavior duration differences, and displacement amplitude differences.

[0057] Input the cognitive task type labels and the behavioral feature difference indicators in the current training stage into the multi-dimensional evaluation model of reverse cognitive transfer interference, and output the cognitive transfer interference value of the animal in the current training stage; Specifically, import the identified key task transition nodes and the corresponding abnormal motor behavior segments in the current training stage into the multi-dimensional evaluation model to evaluate the degree of cognitive transfer interference of the animal in the current training stage.

[0058] Read the cognitive task type label corresponding to the key task conversion node in the current training stage, and match the cognitive task type label with the task switching type classification variable encoding position in the multidimensional evaluation model.

[0059] Behavioral feature difference indicators are extracted from abnormal movement behavior fragments associated with key task transition nodes.

[0060] The current task switching type and behavioral characteristic difference index are input into the multidimensional evaluation model, and the cognitive transfer interference value of the animal in the current training stage is output according to the preset weighted calculation rules.

[0061] Specifically, the cognitive transfer interference degree of the animal in the current training stage is matched with the preset interference threshold, and the sequence arrangement of the cognitive tasks is adjusted according to the matching result, including: Obtaining a preset interference threshold, and comparing the cognitive transfer interference value of the animal in the current training stage with the preset interference threshold; Specifically, through statistical analysis of multiple animal cognitive training experiments, a critical interference value that can represent the upper limit of the normal cognitive transfer fluctuation degree of animals is obtained. This value is defined as a preset interference threshold and is used to determine whether the degree of cognitive transfer interference in the current training stage of the animal requires adjustment of the training task sequence.

[0062] When the cognitive transfer interference value of the animal's current training stage is greater than the preset interference threshold, the cognitive training task sequence is reordered from high to low according to the similarity of cognitive task type labels, and a transition training link is inserted between adjacent cognitive training tasks; Specifically, the cognitive tasks to be performed are reordered according to the cognitive task type labels. The reordering rule is to prioritize placing the training tasks with a higher degree of similarity to the current task type that causes significant cognitive transfer interference at the front of the cognitive training task sequence, and placing the task types with a lower degree of similarity to the current task type at the back of the cognitive training task sequence.

[0063] For example, if the current task causing interference is a path recognition task, recognition tasks similar to the path recognition task will be prioritized in the cognitive training task sequence so that the animals can gradually adapt to similar cognitive patterns; other tasks with large differences in categories, such as spatial jumping tasks, will be placed at the end of the cognitive training task sequence to reduce the frequent and drastic switching of the animals' cognitive patterns.

[0064] A transition training session was inserted between adjacent tasks after reordering. The transition training session was a simplified training task with a cognitive difficulty between the previous and next cognitive tasks, which was used to help the animals gradually adapt to the switching of cognitive modes and reduce the risk of reverse cognitive transfer interference in animals.

[0065] For example, a simple path tracking task is inserted between the path recognition task and the space jump task. Through the training of this path tracking task, the animal gradually adapts to the cognitive transformation from pure visual recognition tasks to action-oriented tasks, reducing the drastic change in cognitive behavior patterns.

[0066] When the cognitive transfer interference value in the current training stage of the animal is less than or equal to the preset interference threshold, the cognitive training task sequence remains unchanged.

[0067] Embodiment 2 The difference between Embodiment 2 and Embodiment 1 of the present invention is that this embodiment introduces a training analysis system for animal cognitive movement.

[0068] Figure 2 The structural schematic diagram of the training analysis system for animal cognitive movement of the present invention is given. The training analysis system for animal cognitive movement includes a behavior acquisition module, a label extraction module, an interference recognition module, an interference evaluation module, and a sequence adjustment module; The behavior acquisition module obtains the original motion behavior data generated by the animal during the cognitive task transformation, performs feature segmentation processing on the original motion behavior data, and generates multiple motion behavior subsequences; The label extraction module extracts the cognitive task labels corresponding to each motion behavior subsequence according to the time nodes of the changes in the cognitive task types received by the animal, and analyzes the behavioral feature differences between adjacent cognitive task subsequences to generate a cross-task difference feature matrix; The interference recognition module establishes a reverse cognitive transfer interference index, identifies abnormal motion behavior segments that deviate from the normal cognitive mode according to the cross-task difference feature matrix, and determines the key task conversion nodes that generate reverse cognitive transfer interference; The interference evaluation module constructs a multi-dimensional evaluation model of reverse cognitive transfer interference based on historical cognitive training data, and evaluates the degree of cognitive transfer interference in the current training stage of the animal according to the key task conversion nodes and the corresponding abnormal motion behavior segments; The sequence adjustment module matches the degree of cognitive transfer interference in the current training stage of the animal with the preset interference threshold, and adjusts the sequence arrangement of the cognitive tasks according to the matching result.

[0069] The above formulas are all dimensionless and take their numerical values for calculation. The formulas are obtained by collecting a large amount of data for software simulation to get a formula that is closest to the real situation. The preset parameters and threshold selection in the formulas are set by those skilled in the art according to the actual situation.

[0070] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the processes or functions described in the embodiments of the present application are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center by wired (such as infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or data center that includes one or more collections of available media. The available media can be magnetic media (such as floppy disks, hard disks, magnetic tapes), optical media (such as DVDs), or semiconductor media. The semiconductor media can be a solid-state drive.

[0071] Those of ordinary skill in the art will realize that the modules and algorithm steps of the examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or in a combination of computer software and electronic hardware. Whether these functions are executed in hardware or software depends on the specific application and design constraints of the technical solution. Skilled artisans can use different methods for each specific application to implement the described functions, but such implementation should not be considered to exceed the scope of this application.

[0072] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the systems, devices, and modules described above can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein.

[0073] In the several embodiments provided in this application, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the modules is only a logical function division, and there can be other division methods in actual implementation. For example, multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed couplings or direct couplings or communication connections to each other can be through some interfaces. The indirect couplings or communication connections of devices or modules can be in electrical, mechanical, or other forms.

[0074] The module described as a separation component may or may not be physically separated. The component shown as a module may or may not be a physical module, and it may be located in one place or distributed across multiple network modules. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0075] In addition, in each embodiment of this application, the various functional modules can be integrated into one processing module, or each module can exist physically alone, or two or more modules can be integrated into one module.

[0076] If the above functions are implemented in the form of software functional modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in each embodiment of this application. The aforementioned storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical discs that can store program codes.

[0077] As described above, this is only the specific implementation manner of this application, but the protection scope of this application is not limited thereto. Any person skilled in the art within the technical scope disclosed in this application can easily think of changes or substitutions, which should all be covered by the protection scope of this application. Therefore, the protection scope of this application should be subject to the protection scope of the claimed rights.

[0078] Finally: The above are only the preferred embodiments of the present invention and are not used to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A training analysis method for animal cognitive movement, characterized in that The steps are as follows: Obtain the original motion behavior data generated by the animal during the cognitive task transition process, perform feature segmentation processing on the original motion behavior data, and generate multiple motion behavior subsequences; According to the time nodes when the animal undergoes changes in the cognitive task type, extract the cognitive task labels corresponding to each motion behavior subsequence, and analyze the behavioral feature differences between adjacent cognitive task subsequences to generate a cross-task difference feature matrix; Establish a reverse cognitive transfer interference index, identify abnormal motion behavior segments that deviate from the normal cognitive pattern based on the cross-task difference feature matrix, and determine the key task transition nodes that generate reverse cognitive transfer interference; Construct a multi-dimensional evaluation model for reverse cognitive transfer interference based on historical cognitive training data, and evaluate the degree of cognitive transfer interference in the current training stage of the animal according to the key task transition nodes and the corresponding abnormal motion behavior segments; Match the degree of cognitive transfer interference in the current training stage of the animal with a preset interference threshold, and adjust the sequence arrangement of cognitive tasks according to the matching result.

2. The training analysis method for animal cognitive movement according to claim 1, wherein, Obtain the original motion behavior data generated by the animal during the cognitive task transition process, perform feature segmentation processing on the original motion behavior data, and generate multiple motion behavior subsequences. Specifically: Perform multi-channel parallel acquisition of the original motion behavior data of the animal during different cognitive task execution stages; Segment the collected original motion behavior data according to the cognitive task trigger time nodes, and divide it into multiple initial behavior paragraph data sets; Based on the initial behavior paragraph data sets, extract the behavioral structure feature indicators in the initial behavior paragraph data to form a structured behavioral feature representation; According to the change pattern of the behavioral structure feature indicators in the structured behavioral feature representation, use a sliding window method for fine-grained segmentation to generate a set of motion behavior subsequences that meet the consistency of a single behavioral feature.

3. The training analysis method for animal cognitive movement according to claim 2, characterized in that, According to the time nodes when the animal undergoes changes in the cognitive task type, extract the cognitive task labels corresponding to each motion behavior subsequence, and analyze the behavioral feature differences between adjacent cognitive task subsequences to generate a cross-task difference feature matrix. Specifically: Match the time range corresponding to each motion behavior subsequence with the time nodes of the cognitive task type change to determine the cognitive task type label corresponding to each motion behavior subsequence; According to the cognitive task type labels of each motion behavior subsequence, define two motion behavior subsequences arranged continuously in the time dimension as a pair of adjacent cognitive task subsequences; For each pair of adjacent cognitive task subsequences, calculate the behavioral feature difference index corresponding to each pair of adjacent cognitive task subsequences; Based on each pair of adjacent cognitive task subsequences and the corresponding behavioral feature difference index, construct a cross-task difference feature matrix.

4. The training analysis method for animal cognitive movement according to claim 3, wherein, Establish a reverse cognitive transfer interference index, identify abnormal motion behavior segments that deviate from the normal cognitive pattern based on the cross-task difference feature matrix, and determine the key task transition nodes that generate reverse cognitive transfer interference. Specifically: Based on the cross-task difference feature matrix, establish the calculation rules for the reverse cognitive transfer interference index; According to the calculation rules of the reverse cognitive transfer interference index, the behavioral feature difference index in the cross-task difference feature matrix is weighted and calculated to output the reverse cognitive transfer interference index value; According to the magnitude relationship between the reverse cognitive transfer interference index value and the preset cognitive mode deviation threshold, it is determined whether the motion behavior subsequence belongs to the motion behavior abnormal segment that deviates from the normal cognitive mode; Based on the adjacent cognitive task subsequences corresponding to the motion behavior abnormal segment, the key task conversion nodes that generate reverse cognitive transfer interference are determined; 5. The training analysis method for animal cognitive movement according to claim 4, wherein A multi-dimensional evaluation model of reverse cognitive transfer interference is constructed based on historical cognitive training data. According to the key task conversion nodes and the corresponding motion behavior abnormal segments, the cognitive transfer interference degree of the animal in the current training stage is evaluated. Specifically: Extract the cognitive task type labels corresponding to the key task conversion nodes and the behavioral feature difference indexes corresponding to the motion behavior abnormal segments during the historical cognitive training process; Based on the influence weights of the cognitive task type labels and the behavioral feature difference indexes on the reverse cognitive transfer interference, a multi-dimensional evaluation model of reverse cognitive transfer interference is generated; Input the cognitive task type labels and the behavioral feature difference indexes in the current training stage into the multi-dimensional evaluation model of reverse cognitive transfer interference, and output the cognitive transfer interference value of the animal in the current training stage; 6. The training analysis method for animal cognitive movement according to claim 5, wherein Match the cognitive transfer interference degree of the animal in the current training stage with the preset interference threshold, and adjust the sequence arrangement of the cognitive tasks according to the matching result. Specifically: Obtain the preset interference threshold, and compare the cognitive transfer interference value of the animal in the current training stage with the preset interference threshold; When the cognitive transfer interference value of the animal in the current training stage is greater than the preset interference threshold, reorder the cognitive training task sequence from high to low according to the similarity of the cognitive task type labels, and insert a transition training link between adjacent cognitive training tasks; When the cognitive transfer interference value of the animal in the current training stage is less than or equal to the preset interference threshold, keep the cognitive training task sequence unchanged; 7. A training analysis system for animal cognitive movement, which is used to implement the training analysis method for animal cognitive movement according to any one of claims 1-6, characterized in that, It includes a behavior acquisition module, a label extraction module, an interference recognition module, an interference evaluation module, and a sequence adjustment module; The behavior acquisition module obtains the original motion behavior data generated by the animal during the cognitive task transition, performs feature segmentation processing on the original motion behavior data, and generates multiple motion behavior subsequences; The label extraction module extracts the cognitive task labels corresponding to each motion behavior subsequence according to the time nodes when the animal receives the changes in the cognitive task types, and analyzes the behavioral feature differences between adjacent cognitive task subsequences to generate a cross-task difference feature matrix; The interference recognition module establishes a reverse cognitive transfer interference index, identifies the motion behavior abnormal segments that deviate from the normal cognitive mode according to the cross-task difference feature matrix, and determines the key task conversion nodes that generate reverse cognitive transfer interference; The interference evaluation module constructs a multi-dimensional evaluation model of reverse cognitive transfer interference based on historical cognitive training data, and evaluates the cognitive transfer interference degree of the animal in the current training stage according to the key task conversion nodes and the corresponding motion behavior abnormal segments; The sequence adjustment module matches the cognitive transfer interference level of the animal in the current training stage with a preset interference threshold, and adjusts the sequence arrangement of the cognitive tasks according to the matching result.

Images