Method for evaluating working stability of guide dogs based on multi-dimensional stimulation environment simulation

Abstract

The application discloses a guide dog working stability evaluation method based on a multi-dimensional stimulation environment simulation and relates to the technical field of guide dog training and evaluation.The application comprises the following steps: S1, constructing a multi-dimensional stimulation test environment and a scene sequence; S2, implementing a standardized test procedure and multi-modal data acquisition; S3, quantitative analysis of stability indexes based on preset observation points; and S4, comprehensive scoring and stability grade determination.The method first systematically integrates multi-dimensional and dynamic environmental stimuli such as social interaction, vision, hearing and sudden fright into a standardized evaluation procedure, highly simulates the most challenging complex scenes that a guide dog may encounter in actual work, and makes the evaluation results have very high ecological validity and predictive value.

Description

Technical Field

[0001] This invention belongs to the field of guide dog training and assessment technology, specifically a method for evaluating the working stability of guide dogs based on multi-dimensional stimulus environment simulation. Background Technology

[0002] Guide dogs are vital partners for visually impaired individuals, ensuring their safe and independent travel. Currently, the selection, training, and assessment of guide dogs primarily focus on "skill mastery" aspects such as basic obedience, route memorization, and obstacle avoidance. Although assessments are conducted in real-world street environments, these are mostly fixed routes. As the lives and work of visually impaired individuals become more diverse, the application scenarios for guide dogs are expanding, constantly raising new requirements for their competence evaluation. In these diverse practical working scenarios, guide dogs face a dynamic, complex, and unpredictable environment, including sudden noises, moving objects, visual disturbances, and the approach of strangers. The existing assessment system still has shortcomings:

[0003] 1. The evaluation dimensions are singular and static: there is a lack of systematic integration and standardized simulation of complex and dynamic environmental stimuli.

[0004] 2. Highly subjective: Relying on the experience and observation of the assessment team members, it is difficult to objectively and quantitatively evaluate the dog's "work stability" under interference, namely its sustained focus, ability to resist interference, and reliability of task execution.

[0005] 3. Limited predictability: The road conditions of the assessment route may vary, which may make it impossible to accurately analyze the long-term working performance and stress threshold of guide dogs in real and complex environments. This may lead to fluctuations in the working status of some dogs after they actually serve, such as distraction, fear, and disobedience to commands.

[0006] Therefore, there is an urgent need to invent a scientific, systematic, and objective evaluation method to conduct a standardized and quantifiable assessment of guide dogs' core working quality—job stability—in a simulated, complex environment. Currently, there is no similar method or system.

[0007] The information disclosed above in this background section is only for enhancing the understanding of the background section of this invention, and therefore may include prior art that is not known to those skilled in the art. Summary of the Invention

[0008] The purpose of this invention is to provide a method for evaluating the working stability of guide dogs based on multi-dimensional stimulus environment simulation, so as to solve the problems in the prior art.

[0009] To achieve the above objectives, the present invention provides the following technical solution: a method for evaluating the working stability of guide dogs based on multi-dimensional stimulus environment simulation, comprising the following steps:

[0010] S1: Constructing a multidimensional stimulus testing environment and scenario sequence:

[0011] Design a physical test space containing a fixed path and multiple stimulus trigger points; pre-set a series of test sub-scenes arranged in logical order or difficulty gradient, each sub-scene simulating one or more specific environmental interference factors;

[0012] S2: Implement standardized testing procedures and multimodal data acquisition:

[0013] The trainer guided the blind dog to complete the test according to the detailed operating procedures of each sub-scenario in the pre-set S1; during the test, the following multimodal data were collected simultaneously:

[0014] (1) Behavioral video data: The dog's body movements, head orientation, gaze focus, movement trajectory, and tail status were recorded throughout the process using fixed and mobile cameras;

[0015] (2) Time node data: Accurately record the time and duration of key events, including the time of instruction issuance, the time of interference, the duration of dog reaction delay, and the total time of task completion;

[0016] (3) Physiological data: Before the start of the test or at a specific resting stage, the resting heart rate of the dog was collected as a baseline reference;

[0017] (4) Trainer Intervention Record: Record the number and methods of additional interventions made by the trainer to soothe or correct the dog's behavior;

[0018] S3: Quantitative analysis of stability indicators based on preset observation points:

[0019] For each test sub-scenario, a set of quantifiable stability metrics are defined, and the collected data is analyzed to calculate the specific stability parameter score.

[0020] S4: Overall Score and Stability Level Determination:

[0021] Weights are assigned to the stability parameters of each sub-scene, and the individual score of the scene is obtained by weighted summation.

[0022] Based on the importance of different sub-scenes in the simulated real work environment, scene weights are set to obtain scene weight scores. According to the mapping relationship between each test sub-scene and the stability dimension, the scene scores are weighted and summed according to the scene weights normalized within the dimension to obtain the sub-scores of the three dimensions of "social environment stability", "sensory interference stability" and "sudden fright stability".

[0023] The scores of each dimension are weighted and summed according to their respective dimensions to obtain the overall work stability score.

[0024] Finally, based on the comprehensive score range, the stability level of the guide dog's work stability is determined.

[0025] Preferably, the test sub-scenarios specifically include:

[0026] A. Social Isolation and Stranger Stress Test: "New Environment Test" assesses the dog's separation anxiety in unfamiliar environments, its persistence in obeying owner commands, and its alertness / attention to passing strangers;

[0027] B. Visual and Dynamic Disturbances Test: "Swinging Doll Test" assesses the dog's level of curiosity, fear, or neglect of sudden, novel, non-threatening visual moving objects and their impact on a given walking task;

[0028] C. Auditory and visual sensory interference tests: “sound test”, “light test” and a combination of the two tests to assess the dog’s ability to adapt to continuous or flashing abnormal sound and light stimuli and its ability to maintain attention;

[0029] D. Simulated tests of sudden fright events: “Sudden appearance of a person”, “Garfield (animated puppet) test” and “Sudden appearance of a sound test” to assess the dog’s stress response intensity, recovery speed and ability to continue to execute commands after task interruption when faced with highly unpredictable and intense stimuli.

[0030] Preferably, the stability parameters include:

[0031] a. Command obedience stability parameters: "Percentage of time spent in a lying position while alone", "Latency of first disobeying command under interference", "Number of times the command needs to be repeated";

[0032] b. Attention allocation stability parameters: "Percentage of time spent looking at the distraction source", "Percentage of time spent looking at the owner or task-related direction", "Frequency of eye contact switching between the distraction source and the task";

[0033] c. Stress response and recovery parameters: "Duration of behavioral rigidity after being startled", "Delay time from the interrupted state to continue performing the task", "Duration or intensity level of abnormal behavior".

[0034] d. Task execution efficiency parameters: "Rate of change in completion time of the same path segment before and after interference", "Path deviation".

[0035] Preferably, the method for quantifying the stability parameter includes:

[0036] a. The instruction conforms to stability parameters:

[0037] (1) "Percentage of time spent in a lying position during solitude" = (cumulative time spent in a lying position / total time spent alone during the test) × 100%. The score is mapped to a percentage range of 0-30 points. ≥90% gets 30 points, 70%-89% gets 20 points, 50%-69% gets 10 points, and <50% gets 0 points.

[0038] (2) "Incubation period of the first disobedience under interference" = the time (seconds) from the occurrence of interference to the first disobedience. The score is graded according to the length of time: >30 seconds gets 35 points, 20-30 seconds gets 25 points, 10-20 seconds gets 15 points, 5-10 seconds gets 5 points, and <5 seconds gets 0 points.

[0039] (3) "Number of times the command needs to be repeated" = the number of times the instructor repeats the same command in a single sub-scene. 5 points will be deducted for each repetition, up to a maximum of 35 points, until all points are deducted.

[0040] b. Attention allocation stability parameters:

[0041] (1) "Percentage of time spent looking at the interference source in total time" = (time spent looking at the interference source / total test time) × 100%. The lower the percentage, the higher the score. The score is assigned in reverse from 0% to 100%, ranging from 0 to 35 points. For every 10% decrease in percentage, the score increases by 3.5 points.

[0042] (2) "Percentage of time spent looking at the owner or the direction related to the task" = (time spent looking at the direction related to the task / total test time) × 100%. The higher the percentage, the higher the score. The score is 0-35 points based on a positive score of 0%-100%. For every 10% increase in percentage, the score increases by 3.5 points.

[0043] (3) "Frequency of eye contact switching" = the number of times the eye contact switches between the interference source and the task per unit time. The lower the frequency, the higher the score. The score is assigned in reverse from 0% to 100% from 0 to 30 points. A frequency of ≤2 times / minute gets 30 points, and 5 points are deducted for each additional 1 time / minute. A frequency of ≥8 times / minute gets 0 points.

[0044] c. Stress response and recovery parameters:

[0045] (1) "Duration of behavioral rigidity" = the time (in seconds) after being startled and stopping the action. The shorter the time, the higher the score. ≤2 seconds get 35 points, 2-5 seconds get 20 points, 5-10 seconds get 10 points, and >10 seconds get 0 points.

[0046] (2) "Recovery delay time" = the time (seconds) from being interrupted to resuming the task. The shorter the time, the higher the score. ≤3 seconds get 35 points, 3-8 seconds get 20 points, 8-15 seconds get 10 points, and >15 seconds get 0 points.

[0047] (3) “Abnormal behavior intensity level” = scored on a scale of 1-5 according to the duration and frequency of barking and hiding behavior: Level 1 (mild): brief barking / hiding, ≤5 seconds, 30 points; Level 2 (mild): lasts 5-15 seconds, 20 points; Level 3 (moderate): lasts 15-30 seconds, 10 points; Level 4 (severe): lasts >30 seconds or occurs multiple times, 5 points; Level 5 (very severe): unable to continue the task, 0 points;

[0048] d. Task execution efficiency parameters:

[0049] (1) "Completion time change rate" = (completion time after interference - completion time before interference) / completion time before interference × 100%. The smaller the change rate, the higher the score. Change rate ≤ 10% gets 50 points, and 10 points are deducted for every 5% increase. ≥ 35% gets 0 points.

[0050] (2) "Path deviation" = the mean square error between the actual walking path and the preset path (meters). The smaller the error, the higher the score; ≤0.5 meters get 50 points, 0.5-0.7 meters get 40 points, 0.7-0.9 meters get 30 points, 0.9-1.1 meters get 20 points, 1.1-1.5 meters get 10 points, and >1.5 meters get 0 points.

[0051] Preferably, the weight settings of the stability parameters are configured differently according to the evaluation objectives of each test sub-scenario, as follows:

[0052] The weighting of each parameter for the social isolation and stranger stress test sub-scenario A is as follows: command obedience stability weight 50%, attention allocation stability weight 20%, stress response and recovery weight 30%;

[0053] The weighting of each parameter for the B-vision and dynamic interference test sub-scene is as follows: instruction obedience stability weight 20%, attention allocation stability weight 50%, and task execution efficiency weight 30%.

[0054] The weighting of each parameter for the C-level auditory and visual sensory interference test sub-scene is as follows: attention allocation stability weight 40%, stress response and recovery weight 30%, and task execution efficiency weight 30%.

[0055] The weights of the parameters for the simulated test sub-scenario of sudden fright event D are as follows: command obedience stability weight 25%, stress response and recovery weight 60%, and task execution efficiency weight 15%.

[0056] Preferably, the scene weight allocation is as follows:

[0057] A social isolation and stranger stress test scenario, weighted at 20%.

[0058] B. Visual and dynamic interference test scenarios have a weighting of 25%.

[0059] C. Auditory and visual sensory interference test scenario weighting: 30%

[0060] D. Sudden frightening event simulation test scenario weight 25%.

[0061] Preferably, the normalized scene weights within the dimension are allocated as follows:

[0062] "Social environment stability" only includes the "A social isolation and stranger stress test" scenario, with a weight of 100% within this dimension;

[0063] "Sensory interference stability" includes "B visual and dynamic interference test" and "C auditory and visual sensory interference test" scenarios. The normalized weight within the dimension is 45.45% for scenario B and 54.55% for scenario C.

[0064] "Sudden fright stability" only includes the "D sudden fright event simulation test" scenario, with a weight of 100% within the dimension.

[0065] Preferably, the dimension weights are allocated as follows: social environment stability 20%, sensory disturbance stability 35%, and sudden fright stability 45%.

[0066] Preferably, the stability levels include excellent, good, acceptable, requires reinforcement, and unacceptable, and the correspondence between the comprehensive score range and the stability levels is as follows:

[0067] Overall score ≥ 90 points: Excellent

[0068] 80 points ≤ Overall score < 90 points: Good

[0069] 70 points ≤ Overall score < 80 points: Pass

[0070] 60 points ≤ Overall score < 70 points: Needs reinforcement.

[0071] Overall score < 60 points: unqualified.

[0072] The guide dog working stability assessment system based on multi-dimensional stimulus environment simulation, as described above, includes:

[0073] The test environment control unit is used to configure and trigger stimuli in each test sub-scene, including sound, light, dynamic puppets, and the appearance of strangers.

[0074] The multimodal data acquisition unit includes:

[0075] The video capture module is equipped with multiple high-definition cameras to capture the dog's behavior, gaze, and trajectory.

[0076] Physiological signal acquisition module, wearable heart rate monitoring device, collects heart rate data in real time;

[0077] The time synchronization module is used to mark timestamps for key events such as instructions, interference, and reactions.

[0078] The trainer interaction recording module is used to record the number of times the trainer gives commands and their intervention behaviors;

[0079] The data processing and storage unit is used to receive, store, and preprocess the acquired raw data, and to perform time alignment and format standardization.

[0080] The stability index quantitative analysis unit automatically calculates the parameter scores for each sub-scenario based on a preset stability parameter algorithm.

[0081] The weight configuration and management unit is used to set and adjust parameter weights, scene weights, and dimension weights.

[0082] The comprehensive scoring calculation unit calculates the sub-scene score, dimension score, and comprehensive score by weighting them layer by layer according to the weight configuration.

[0083] The stability level determination unit automatically outputs the stability level and evaluation report based on the comprehensive score range.

[0084] The system management and human-computer interaction interface provides a user interface for test process control, data viewing, parameter configuration, and report export.

[0085] Compared with the prior art, the present invention has the following beneficial effects:

[0086] 1. The method of this invention is the first to systematically integrate multi-dimensional and dynamic environmental stimuli such as social, visual, auditory, and sudden fright into a standardized assessment process, which highly simulates the most challenging and complex scenarios that guide dogs may encounter in actual work, giving the assessment results extremely high ecological validity and predictive value.

[0087] 2. The method of this invention completely changes the evaluation model that relies on the subjective impressions of the assessment team members by defining precise behavioral observation points such as the percentage of eye contact time, reaction latency, and task time difference, and converting them into calculable parameters. This makes the evaluation results measurable, comparable, and traceable, greatly improving the scientific nature and fairness of the assessment.

[0088] 3. The method of this invention can not only give an overall conclusion on whether the dog is qualified or not, but also accurately diagnose the specific weaknesses in the dog's working stability through multi-dimensional sub-item scores. For example, whether it is difficult to adapt to continuous noise or oversensitivity to sudden visual stimuli. This provides precise targeted guidance for subsequent personalized reinforcement training and greatly improves training efficiency.

[0089] 4. The method of this invention provides guide dog training institutions and certification organizations with a set of standardized quality control tools that can be unified nationwide and even globally, ensuring the minimum guarantee standard of working performance of guide dogs in service and reducing the risks caused by the instability of the dog's condition.

[0090] 5. All high-intensity stimuli in this invention are conducted in a controlled simulated environment, avoiding the safety risks that may arise from real street testing, while also protecting the physical and mental health of the trained dogs and allowing them to explore stress thresholds within a safe range.

[0091] 6. Applying this assessment method to the early evaluation of guide dog puppies or prospective breeding dogs can identify individuals with innate advantages in stability traits, providing a key behavioral screening tool for scientific and efficient guide dog bloodline breeding, and improving the overall working quality of the dog population from the source.

[0092] 7. The core principle of the method of this invention can be adapted and applied to the assessment of stress resistance, environmental adaptability and task focus of other types of working dogs such as police dogs, search and rescue dogs, and medical testing dogs, and has broad market expansion potential. Attached Figure Description

[0093] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0094] Figure 1 This is a schematic diagram of the social isolation and stranger stress test scenario of the present invention. Figure 1 ;

[0095] Figure 2 This is a schematic diagram of the social isolation and stranger stress test scenario of the present invention. Figure 2 ;

[0096] Figure 3 This is a schematic diagram of the visual and dynamic interference test scenario of the present invention;

[0097] Figure 4 This is a schematic diagram of the auditory and visual sensory interference test scenario of the present invention;

[0098] Figure 5 This is a schematic diagram of a simulated test scenario for a sudden frightening event according to the present invention;

[0099] Figure 6 This is a schematic diagram of the "sudden sound test" scenario of the present invention. Detailed Implementation

[0100] To enable those skilled in the art to better understand the technical solutions of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0101] Example 1: Evaluation process of the working stability of guide dog "Dodo"

[0102] I. Calculation of Measured Data and Parameter Scores for Each Sub-Scene

[0103] A. Social Distancing and Stranger Stress Test

[0104] Test content: The dog is left alone in an unfamiliar environment, during which a stranger passes by and approaches.

[0105] Actual measurement data:

[0106] 1. Instructions conform to stability parameters:

[0107] (1) Total time spent in a lying position while alone: ​​180 seconds, total test time: 200 seconds;

[0108] Percentage of time spent in a lying position while alone = (180 / 200) × 100% = 90%, 30 points;

[0109] (2) Incubation period for the first disobedience under interference: 25 seconds, after the appearance of the stranger;

[0110] 25 points in 20-30 seconds;

[0111] (3) Number of times the instructor repeats the command: 2 times

[0112] The maximum score is 35 points. Repeating the command twice will deduct 10 points, and the score will be 25 points.

[0113] Subtotal: 30 + 25 + 25 = 80 points.

[0114] 2. Attention allocation stability parameters:

[0115] (1) Percentage of time spent staring at strangers: 15%;

[0116] 35 × (1 - 0.15) = 29.75 points;

[0117] (2) Percentage of time spent focusing on the task direction: 70%;

[0118] 35 × 0.7 = 24.5 points;

[0119] (3) Frequency of eye contact switching: 4 times / minute;

[0120] 30 - 5 × (4 - 2) = 20 points;

[0121] Subtotal: 29.75 + 24.5 + 20 = 74.25 points.

[0122] 3. Stress response and recovery parameters:

[0123] (1) Duration of behavioral stiffness after being startled: 3 seconds, when a stranger suddenly approaches;

[0124] 20 points are awarded for 2-5 seconds;

[0125] (2) Delay time from the interrupted state to the continuation of the task: 9 seconds;

[0126] 10 points are awarded in 8-15 seconds;

[0127] (3) Abnormal behavior intensity level: Level 1, slight barking, lasting 4 seconds.

[0128] 30 points;

[0129] Subtotal: 20 + 10 + 30 = 60 points.

[0130] Score for Scenario A: 80×0.5+74.25×0.2+60×0.3=72.85 points.

[0131] B. Visual and Dynamic Interference Test

[0132] Test content: First, lie down and wait, then encounter a rocking puppet while moving.

[0133] Actual measurement data:

[0134] 1. Instructions conform to stability parameters:

[0135] (1) Percentage of time spent in a lying position while alone: ​​100%, wait 30 seconds to remain in a lying position throughout;

[0136] 30 points;

[0137] (2) Latency period for the first disobedience under interference: 15 seconds;

[0138] 15 points are awarded in 10-20 seconds;

[0139] (3) Number of times the password needs to be repeated: 1 time;

[0140] Each repetition deducts 5 points, resulting in a total of 30 points.

[0141] Subtotal: 30 + 15 + 30 = 75 points.

[0142] 2. Attention allocation stability parameters:

[0143] (1) Percentage of time spent staring at the doll: 20%;

[0144] 35 × (1 - 0.2) = 28 points;

[0145] (2) Percentage of time spent focusing on the task direction: 65%;

[0146] 35 × 0.65 = 22.75 points;

[0147] (3) Frequency of eye contact switching: 5 times / minute;

[0148] 30 - 5 × (5 - 2) = 15 points;

[0149] Subtotal: 28 + 22.75 + 15 = 65.75 points.

[0150] 3. Task execution efficiency parameters:

[0151] (1) Rate of change in completion time: 12%;

[0152] 50 - 10 × ((12 - 10) / 5) = 46 points;

[0153] (2) Path deviation: 0.6 meters;

[0154] 0.5-0.7 meters scores 40 points.

[0155] Subtotal: 46 + 40 = 86 points.

[0156] Score for Scenario B: 75×0.2+65.75×0.5+86×0.3=73.675 points.

[0157] C. Auditory and visual sensory interference test

[0158] Test content: Proceeding under continuous sound and flashing light interference.

[0159] Actual measurement data:

[0160] 1. Attention allocation stability parameters:

[0161] (1) Percentage of time spent looking at the source of interference: 25%;

[0162] 35 × (1 - 0.25) = 26.25 points;

[0163] (2) Percentage of time spent focusing on the task direction: 60%;

[0164] 35 × 0.6 = 21 points;

[0165] (3) Frequency of eye contact switching: 6 times / minute;

[0166] 30 - 5 × (6 - 2) = 10 points;

[0167] Subtotal: 26.25 + 21 + 10 = 57.25 points.

[0168] 2. Stress response and recovery parameters:

[0169] (1) Duration of behavioral rigidity: 4 seconds;

[0170] 20 points are awarded for 2-5 seconds;

[0171] (2) Recovery delay time: 10 seconds;

[0172] 10 points are awarded in 8-15 seconds;

[0173] (3) Intensity of abnormal behavior: Level 2, mild, lasting 10 seconds;

[0174] 20 points;

[0175] Subtotal: 20 + 10 + 20 = 50 points.

[0176] 3. Task execution efficiency parameters:

[0177] (1) Rate of change in completion time: 18%;

[0178] 50 - 10 × ((18 - 10) / 5) = 34 points;

[0179] (2) Path deviation: 0.8 meters;

[0180] 30 points;

[0181] Subtotal: 34 + 30 = 64 points.

[0182] Score for scenario C: 57.25×0.4+50×0.3+64×0.3=57.1 points.

[0183] D. Simulation Test of Sudden Startling Events

[0184] Test content: A moving puppet (Garfield) suddenly appears while walking.

[0185] Actual measurement data:

[0186] 1. Instructions conform to stability parameters:

[0187] (1) Percentage of time spent in a lying position while alone: ​​100%, with a brief 10-second wait in a lying position throughout;

[0188] 30 points;

[0189] (2) Latency period for the first disobedience under interference: 8 seconds;

[0190] 5 points are awarded for 5-10 seconds;

[0191] (3) Number of times the password needs to be repeated: 1 time;

[0192] Each repetition deducts 5 points, resulting in a total of 30 points.

[0193] Subtotal: 30 + 5 + 30 = 65 points.

[0194] 2. Stress response and recovery parameters:

[0195] (1) Duration of behavioral rigidity: 6 seconds;

[0196] 10 points are awarded in 5-10 seconds;

[0197] (2) Recovery delay time: 12 seconds;

[0198] 10 points are awarded in 8-15 seconds;

[0199] (3) Intensity of abnormal behavior: Level 3, moderate, lasting 20 seconds;

[0200] 10 points;

[0201] Subtotal: 10 + 10 + 10 = 30 points.

[0202] 3. Task execution efficiency parameters:

[0203] (1) Rate of change in completion time: 25%;

[0204] 50 - 10 × ((25 - 10) / 5) = 20 points;

[0205] (2) Path deviation: 1.0 meter;

[0206] 20 points;

[0207] Subtotal: 20 + 20 = 40 points.

[0208] Score for Scene D: 65×0.25+30×0.6+40×0.15=40.25 points.

[0209] II. Dimensional Score Calculation

[0210] 1. Social environment stability dimension, including only scenario A, with a weight of 100%;

[0211] Dimensional score = 72.85 points.

[0212] 2. Sensory interference stability dimension, including scenarios B and C, with normalized weights of B: 45.45% and C: 54.55%;

[0213] Dimensional score = 73.675 × 0.4545 + 57.1 × 0.5455 = 33.49 + 31.14 = 64.63 points.

[0214] 3. Stability dimension for sudden fright: Includes only scenario D, weight 100%;

[0215] Dimensional score = 40.25 points.

[0216] III. Overall Score

[0217] Social environment stability: 72.85 × 0.2 = 14.57;

[0218] Sensory interference stability: 64.63 × 0.35 = 22.6205;

[0219] Stability under sudden fright: 40.25 × 0.45 = 18.1125;

[0220] Overall score = 14.57 + 22.6205 + 18.1125 = 55.303 points.

[0221] IV. Stability Level Determination

[0222] Based on the overall score range: Overall score < 60 points: unqualified.

[0223] V. Evaluation Report

[0224] Evaluation results: The guide dog "Duoduo" scored 55.3 points in overall work stability, and was rated as "unqualified".

[0225] Diagnostic recommendations:

[0226] 1. To address the lack of stability in the face of sudden fright, conduct systematic desensitization training, gradually introduce controllable sudden stimuli, and strengthen recovery training.

[0227] 2. Strengthen attention maintenance training under sensory interference, such as continuous task execution under mixed sound and light interference.

[0228] 3. Increase the diversity of interactions with strangers in social settings to enhance resilience and the persistence of obedience.

[0229] 4. It is recommended to reassess after the training cycle. Only those with a comprehensive score of 70 or above can proceed to the next stage of evaluation.

[0230] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A method for evaluating the working stability of guide dogs based on multi-dimensional stimulus environment simulation, characterized by: Includes the following steps: S1: Constructing a multidimensional stimulus testing environment and scenario sequence: Design a physical test space containing a fixed path and multiple stimulus trigger points; pre-set a series of test sub-scenes arranged in logical order or difficulty gradient, each sub-scene simulating one or more specific environmental interference factors; S2: Implement standardized testing procedures and multimodal data acquisition: The trainer led the blind dog to complete the test according to the detailed operation procedures of each sub-scenario in the pre-set S1; during the test, the following multimodal data were collected simultaneously: behavioral video data, time point data, physiological index data, and trainer intervention records; S3: Quantitative analysis of stability indicators based on preset observation points: For each test sub-scenario, a set of quantifiable stability metrics are defined, and the collected data is analyzed to calculate the specific stability parameter score. S4: Overall Score and Stability Level Determination: Weights are assigned to the stability parameters of each sub-scene, and the individual score of the scene is obtained by weighted summation. Based on the importance of different sub-scenes in the simulated real work environment, scene weights are set to obtain scene weight scores. According to the mapping relationship between each test sub-scene and the stability dimension, the scene scores are weighted and summed according to the scene weights normalized within the dimension to obtain the sub-scores of the three dimensions of "social environment stability", "sensory interference stability" and "sudden fright stability". The scores of each dimension are weighted and summed according to their respective dimensions to obtain the overall work stability score. Finally, based on the comprehensive score range, the stability level of the guide dog's work stability is determined.

2. The method for evaluating the working stability of guide dogs based on multi-dimensional stimulus environment simulation according to claim 1, characterized in that: The test sub-scenarios specifically include: A. Social Isolation and Stranger Stress Test: "New Environment Test" assesses the dog's separation anxiety in unfamiliar environments, persistence of obedience to owner commands, and alertness / attention to passing strangers; B. Visual and Dynamic Disturbances Test: "Swinging Doll Test" assesses the dog's level of curiosity, fear, or neglect of sudden, novel, non-threatening visual moving objects and their impact on a given walking task; C. Auditory and visual sensory interference tests: "Sound test", "Light test" and a combination of the two tests, to assess the dog's ability to adapt to continuous or flashing abnormal sound and light stimuli and its ability to maintain attention; D. Simulated tests of sudden fright events: "Sudden appearance of a person", "Garfield (animated puppet) test" and "Sudden appearance of a sound test" to assess the dog's stress response intensity, recovery speed and ability to continue executing commands after task interruption when faced with highly unpredictable and intense stimuli.

3. The method for evaluating the working stability of guide dogs based on multi-dimensional stimulus environment simulation according to claim 2, characterized in that: The stability parameters include: a. Command compliance stability parameters: "Percentage of time spent in a lying position while alone", "Latency of first command disobedience under interference", "Number of times the command needs to be repeated"; b. Attention allocation stability parameters: "Percentage of time spent looking at the distraction source out of the total time", "Percentage of time spent looking at the direction related to the master or task", "Frequency of eye contact switching between the distraction source and the task"; c. Stress response and recovery parameters: "Duration of behavioral rigidity after being startled", "Delay time from the interrupted state to continue performing the task", "Duration or intensity level of abnormal behavior"; d. Task execution efficiency parameters: "rate of change in completion time of the same path segment before and after interference", "path deviation".

4. The method for evaluating the working stability of guide dogs based on multi-dimensional stimulus environment simulation according to claim 3, characterized in that: The methods for quantifying the stability parameters include: a. The instruction conforms to stability parameters: (1) "Percentage of time spent in a lying position during solitude" = (cumulative time spent in a lying position / total time spent alone during the test) × 100%, and the score is mapped to a percentage range of 0-30 points; (2) "Incubation period of the first disobedience under interference" = the time (seconds) from the occurrence of interference to the first disobedience, graded by time length from 0 to 35 points; (3) "Number of times the command needs to be repeated" = the number of times the instructor repeats the same command in a single sub-scene. 5 points will be deducted for each repetition, up to a maximum of 35 points, until all points are deducted. b. Attention allocation stability parameters: (1) "Percentage of time spent looking at the interference source in total time" = (time spent looking at the interference source / total test time) × 100%. The lower the percentage, the higher the score. Scores are assigned in reverse from 0% to 100%, ranging from 0 to 35 points. (2) "Percentage of time spent looking at the owner or in the direction of the task" = (time spent looking at the direction of the task / total test time) × 100%. The higher the percentage, the higher the score. The score is 0-35 points based on a positive score of 0%-100%. (3) "Frequency of eye contact switching" = the number of times the eye contact switches between the source of interference and the task per unit time. The lower the frequency, the higher the score. The score is assigned in reverse from 0% to 100% and is 0-30 points. c. Stress response and recovery parameters: (1) "Duration of behavioral rigidity" = the time (in seconds) after being startled and stopping the action. The shorter the time, the higher the score, 0-35 points; (2) "Recovery delay time" = the time (in seconds) from the interruption to the resumption of the task. The shorter the time, the higher the score, 0-35 points; (3) "Abnormal behavior intensity level" = scored on a scale of 1-5, 0-30 points, based on the duration and frequency of barking and hiding behaviors; d. Task execution efficiency parameters: (1) "Completion time change rate" = (Completion time after interference - Completion time before interference) / Completion time before interference × 100%. The smaller the change rate, the higher the score, 0-50 points; (2) "Path deviation" = the mean square error (meters) between the actual walking path and the preset path. The smaller the error, the higher the score, 0-50 points.

5. The method for evaluating the working stability of guide dogs based on multi-dimensional stimulus environment simulation according to claim 3, characterized in that: The weights of the stability parameters are configured differently based on the evaluation objectives of each test sub-scenario, as follows: The weighting of each parameter for the social isolation and stranger stress test sub-scenario A is as follows: command obedience stability weight 50%, attention allocation stability weight 20%, stress response and recovery weight 30%; The weighting of each parameter for the B-vision and dynamic interference test sub-scene is as follows: instruction obedience stability weight 20%, attention allocation stability weight 50%, and task execution efficiency weight 30%. The weighting of each parameter for the C-level auditory and visual sensory interference test sub-scene is as follows: attention allocation stability weight 40%, stress response and recovery weight 30%, and task execution efficiency weight 30%. The weights of the parameters for the simulated test sub-scenario of sudden fright event D are as follows: command obedience stability weight 25%, stress response and recovery weight 60%, and task execution efficiency weight 15%.

6. The method for evaluating the working stability of guide dogs based on multi-dimensional stimulus environment simulation according to claim 2, characterized in that: The scene weights are assigned as follows: A social isolation and stranger stress test scenario, weighted at 20%. B. Visual and dynamic interference test scenarios have a weighting of 25%. C. Auditory and visual sensory interference test scenario weighting: 30% D. Sudden frightening event simulation test scenario weight 25%.

7. The method for evaluating the working stability of guide dogs based on multi-dimensional stimulus environment simulation according to claim 2, characterized in that: The normalized scene weights within the dimension are assigned as follows: "Social environment stability" only includes the "A social isolation and stranger stress test" scenario, with a weight of 100% within this dimension; "Sensory interference stability" includes "B visual and dynamic interference test" and "C auditory and visual sensory interference test" scenarios. The normalized weight within the dimension is 45.45% for scenario B and 54.55% for scenario C. "Sudden fright stability" only includes the "D sudden fright event simulation test" scenario, with a weight of 100% within the dimension.

8. The method for evaluating the working stability of guide dogs based on multi-dimensional stimulus environment simulation according to claim 1, characterized in that: The weights for the dimensions are as follows: social environment stability 20%, sensory disturbance stability 35%, and sudden fright stability 45%.

9. The method for evaluating the working stability of guide dogs based on multi-dimensional stimulus environment simulation according to claim 1, characterized in that: The stability levels include Excellent, Good, Satisfactory, Needs Improvement, and Unsatisfactory. The correspondence between the comprehensive score range and the stability levels is as follows: Overall score ≥ 90 points: Excellent 80 points ≤ Overall score < 90 points: Good 70 points ≤ Overall score < 80 points: Pass 60 points ≤ Overall score < 70 points: Needs reinforcement. Overall score < 60 points: unqualified.

10. The guide dog working stability evaluation system based on multi-dimensional stimulus environment simulation according to any one of claims 1-9, characterized in that, include: The test environment control unit is used to configure and trigger stimuli in each test sub-scene, including sound, light, dynamic puppets, and the appearance of strangers. The multimodal data acquisition unit includes: The video capture module is equipped with multiple high-definition cameras to capture the dog's behavior, gaze, and trajectory. Physiological signal acquisition module, wearable heart rate monitoring device, collects heart rate data in real time; The time synchronization module is used to mark timestamps for key events such as instructions, interference, and reactions. The trainer interaction recording module is used to record the number of times the trainer gives commands and their intervention behaviors; The data processing and storage unit is used to receive, store, and preprocess the acquired raw data, and to perform time alignment and format standardization. The stability index quantitative analysis unit automatically calculates the parameter scores for each sub-scenario based on a preset stability parameter algorithm. The weight configuration and management unit is used to set and adjust parameter weights, scene weights, and dimension weights. The comprehensive scoring calculation unit calculates the sub-scene score, dimension score, and comprehensive score by weighting them layer by layer according to the weight configuration. The stability level determination unit automatically outputs the stability level and evaluation report based on the comprehensive score range. The system management and human-computer interaction interface provides a user interface for test process control, data viewing, parameter configuration, and report export.

Images