A humanoid robot multi-floor navigation capability testing method, system and device
By designing specialized test tasks and constructing a three-dimensional evaluation system, the shortcomings in evaluating the multi-floor navigation capabilities of humanoid robots were addressed. This enabled the scientific quantification of vertical movement, floor recognition, and conversion capabilities, and enhanced the relevance of the test results to practical application scenarios.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2025-06-20
- Publication Date
- 2026-04-14
AI Technical Summary
Existing technologies lack a systematic evaluation of the multi-floor navigation capabilities of humanoid robots, especially in terms of vertical movement capabilities, floor recognition and transition capabilities, positioning accuracy in multi-floor environments, and navigation continuity during floor transitions.
Design specific test tasks such as stair climbing and elevator use, combine quantitative indicators such as floor recognition accuracy and navigation performance, construct a three-dimensional evaluation system, and construct a weighted evaluation model through indicators such as positioning continuity and the rationality of floor conversion strategies to achieve scientific quantitative evaluation in multi-story building environments.
It significantly enhances the fit between test results and actual application scenarios, provides accurate data to support algorithm optimization and hardware upgrades, solves the problems of inaccurate positioning accuracy and poor navigation continuity during floor transitions in multi-story environments, and realizes the scientific quantification and horizontal comparability of robot navigation capabilities.
Smart Images
Figure CN120369006B_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of robotics technology, and in particular to a method, system and equipment for testing the multi-floor navigation capability of a humanoid robot. Background Technology
[0002] With the rapid development of artificial intelligence and robotics, humanoid robots have been widely used in various scenarios such as home services, commercial services, medical assistance, and security patrols. They often work in multi-story building environments, which requires them to have good multi-floor navigation capabilities.
[0003] Currently, humanoid robot performance testing mainly focuses on motion, obstacle avoidance, and navigation accuracy in single-story planar environments. Related standard documents also primarily provide testing methods for single-story planar environments, lacking a systematic evaluation of navigation capabilities in multi-story building environments. Existing robot navigation testing methods have many shortcomings, such as a lack of evaluation of vertical movement capabilities, floor recognition and transition capabilities, positioning accuracy in multi-story environments, and navigation continuity during floor transitions. These methods cannot comprehensively and scientifically verify the navigation performance of humanoid robots in multi-story building environments. Summary of the Invention
[0004] This application provides a method, system, and device for testing the multi-floor navigation capability of a humanoid robot, in order to solve the problems in the prior art such as the lack of vertical movement capability assessment, the inability to assess floor recognition and transition capabilities, inaccurate positioning accuracy in multi-floor environments, and poor continuity of floor transition navigation.
[0005] The first aspect of this application provides a method for testing the multi-floor navigation capability of a humanoid robot, comprising the following steps: selecting a target test environment; selecting no fewer than 5 reference points on each floor, instructing the robot to move to each point, recording the robot's self-localization coordinates and the measured coordinates of the motion capture system, and calculating the localization error; randomly placing the robot on different floors, instructing it to identify the current floor, recording the identification results, and calculating the floor identification accuracy and average identification time; setting up a cross-floor staircase navigation task, requiring the robot to travel from one floor to another via stairs, and evaluating the staircase navigation performance; setting up a cross-floor elevator navigation task, requiring the robot to travel from one floor to another via elevator, and evaluating the elevator navigation performance; setting up a complex multi-floor navigation task, including multiple target points and floor transitions, and evaluating the comprehensive navigation capability;
[0006] A weighted evaluation model is constructed based on the positioning error, the floor recognition accuracy, the stair navigation performance, the elevator navigation performance, and the overall navigation capability. A comprehensive score is calculated based on each indicator, and a test report is generated.
[0007] Optionally, the stair navigation performance includes stair recognition success rate, climbing speed, posture stability, and positioning continuity; the elevator navigation performance includes elevator recognition success rate, call success rate, entry accuracy, and total navigation time; and the comprehensive navigation capability includes task completion rate, target point positioning accuracy, path optimization rate, and the rationality of floor transition strategies.
[0008] Optionally, the posture stability is the maximum deviation of the torso tilt angle during climbing, the positioning continuity is the jump value of the positioning error when changing floors, and the rationality of the floor changing strategy is the percentage of times that the elevator / stairs is selected in accordance with the requirements of the scenario.
[0009] Optionally, the target testing environment is a standard testing building with at least two floors, and a high-precision motion capture system and standardized reference points are set up on each floor.
[0010] Optionally, the high-precision motion capture system includes at least 8 infrared cameras covering the test area on each floor, with a measurement accuracy better than 1mm; the standardized reference points are no less than 10 on each floor, distributed in offices, corridors, and stairwells, and their three-dimensional coordinates are measured using a total station.
[0011] Optionally, the formula for calculating the positioning error is:
[0012]
[0013] in, Use it as a coordinate system for self-positioning. These are the measured coordinates.
[0014] Optionally, the formula for calculating the overall score based on the various indicators is as follows:
[0015] Overall score = Basic positioning performance score * 20% + Floor recognition ability score * 15% + Staircase navigation ability score * 25% + Elevator navigation ability score * 25% + Overall navigation ability score * 15%.
[0016] Optionally, it also includes: comparing and analyzing the test report with human navigation data under the same task, with comparison dimensions including navigation time, path optimization rate, and rationality of floor transition decisions.
[0017] A second aspect of this application provides a multi-floor navigation capability testing system for a humanoid robot, comprising: an environment configuration module for configuring the test building, deploying a motion capture system and standardized reference points; a positioning test module for performing basic positioning accuracy tests and calculating positioning errors; a floor recognition module for evaluating the robot's floor recognition capability and environmental adaptability; a stair navigation module for testing the robot's stair navigation performance and generating evaluation indicators; an elevator navigation module for testing the robot's elevator navigation performance and generating evaluation indicators; a comprehensive task module for performing multi-floor comprehensive navigation tasks and evaluating comprehensive performance; and a data analysis module for processing test data, calculating comprehensive scores, and generating test reports.
[0018] A third aspect of this application provides an electronic device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement a method for testing the multi-floor navigation capability of a humanoid robot as described in the above embodiments.
[0019] Therefore, this application has at least the following beneficial effects:
[0020] This application's embodiments, through the design of specialized test tasks such as stair climbing and elevator use, combined with quantitative indicators such as floor recognition accuracy and navigation performance, extend the evaluation dimension from a two-dimensional plane to a three-dimensional space, filling the technical gap in the evaluation of vertical movement and dynamic floor transition capabilities. Simultaneously, by utilizing indicators such as positioning continuity and the rationality of floor transition strategies, a three-dimensional evaluation system covering the entire process of multi-story building navigation is constructed, significantly enhancing the fit between test results and actual application scenarios. Furthermore, by accurately calculating positioning errors and linking them with the target point positioning accuracy indicators in the comprehensive navigation task, the impact of complex environments on robot positioning is analyzed in depth, providing precise data support for algorithm optimization and hardware upgrades. The constructed weighted evaluation model incorporates core performance indicators into the comprehensive score calculation, forming a standardized evaluation paradigm, enabling the scientific quantification and horizontal comparability of navigation capabilities of different robots in multi-story building environments. Thus, it solves the problems in existing technologies such as the lack of vertical movement capability evaluation, the inability to evaluate floor recognition and transition capabilities, inaccurate positioning accuracy in multi-story environments, and poor continuity of floor transition navigation.
[0021] Additional aspects and advantages of this application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of this application. Attached Figure Description
[0022] The above and / or additional aspects and advantages of this application will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein:
[0023] Figure 1This is a flowchart illustrating a method for testing the multi-floor navigation capability of a humanoid robot according to an embodiment of this application;
[0024] Figure 2 This is a schematic diagram of a test environment provided according to an embodiment of this application;
[0025] Figure 3 This is a schematic diagram of a stair navigation test scenario provided according to an embodiment of this application;
[0026] Figure 4 This is a schematic diagram of an elevator navigation test scenario according to an embodiment of this application;
[0027] Figure 5 This is an example diagram of an integrated navigation task provided according to an embodiment of this application;
[0028] Figure 6 This is a flowchart of test data processing according to an embodiment of this application;
[0029] Figure 7 This is a system architecture diagram of the evaluation index provided according to an embodiment of this application;
[0030] Figure 8 This is a flowchart illustrating a method for testing the multi-floor navigation capability of a humanoid robot according to an embodiment of this application;
[0031] Figure 9 This is a block diagram illustrating a multi-floor navigation capability testing system for a humanoid robot according to an embodiment of this application.
[0032] Figure 10 This is a schematic diagram of the structure of an electronic device provided according to an embodiment of this application. Detailed Implementation
[0033] The embodiments of this application are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain this application, and should not be construed as limiting this application.
[0034] The following description, with reference to the accompanying drawings, illustrates a method, system, and device for testing the multi-floor navigation capabilities of a humanoid robot according to embodiments of this application. Addressing the problem of poor navigation continuity during floor transitions mentioned in the background section, this application provides a method for testing the multi-floor navigation capabilities of a humanoid robot. This method designs specialized test tasks such as stair climbing and elevator use, and combines quantitative indicators such as floor recognition accuracy and navigation performance to extend the evaluation dimension from a two-dimensional plane to a three-dimensional space, filling the technical gap in evaluating vertical movement and dynamic floor transition capabilities. Simultaneously, by utilizing indicators such as positioning continuity and the rationality of floor transition strategies, a three-dimensional evaluation system covering the entire navigation process in multi-story buildings is constructed, significantly enhancing the consistency between test results and actual application scenarios. Furthermore, by accurately calculating positioning errors and linking them with the target point positioning accuracy indicators in the comprehensive navigation task, the impact of complex environments on robot positioning is analyzed in depth, providing precise data support for algorithm optimization and hardware upgrades. The constructed weighted evaluation model incorporates core performance indicators into the comprehensive score calculation, forming a standardized evaluation paradigm and enabling the scientific quantification and horizontal comparability of the navigation capabilities of different robots in multi-story building environments. This solves the problems in existing technologies, such as the lack of vertical mobility assessment, the inability to assess floor identification and transition capabilities, inaccurate positioning accuracy in multi-floor environments, and poor continuity of floor transition navigation.
[0035] The following description, with reference to the accompanying drawings, describes a method, system, and device for testing the multi-floor navigation capability of a humanoid robot according to embodiments of this application.
[0036] Specifically, Figure 1 This is a flowchart illustrating a method for testing the multi-floor navigation capability of a humanoid robot, as provided in an embodiment of this application.
[0037] like Figure 1 As shown, the method for testing the multi-floor navigation capability of this humanoid robot includes the following steps:
[0038] In step S1, select the target test environment.
[0039] The target testing environment is a standard test building with at least two floors, with a high-precision motion capture system and standardized reference points set up on each floor.
[0040] It is understood that the target test environment selected in this application embodiment is to facilitate the construction of a physical or virtual space that is highly consistent with the actual application scenario of the robot. By pre-setting multi-story building structures, typical environmental elements (such as stairs, elevators, obstacles) and sensor adaptation conditions, a standardized and reproducible evaluation benchmark is provided for navigation performance testing.
[0041] It should be noted that the high-precision motion capture system includes at least 8 infrared cameras covering the test area on each floor, with a measurement accuracy better than 1mm; there are no fewer than 10 standardized reference points on each floor, distributed in offices, corridors, and stairwells, and the three-dimensional coordinates are measured using a total station.
[0042] Specifically, such as Figure 2 As shown, select a standard test building with at least two floors, including staircases, elevators, and various functional areas (such as offices, meeting rooms, corridors, etc.). Each floor of the building should have an area of no less than 200 square meters and a floor height of no less than 3 meters. The floors should be connected by standard staircases (step height 15-18 cm, width 28-32 cm) and standard elevators (door width no less than 80 cm, car depth no less than 120 cm).
[0043] High-precision motion capture systems (such as Vicon and OptiTrack) are installed on each floor of the test building. Each system includes at least eight infrared cameras, covering the entire test area, ensuring measurement accuracy better than 1 millimeter. The coordinate system of the motion capture systems needs to be standardized to facilitate cross-floor position comparison.
[0044] Standardized reference points should be set up within the test building, with no fewer than 10 on each floor. These reference points should be distributed across different functional areas and corridors. The three-dimensional coordinates of each reference point should be accurately measured and recorded using a total station or other high-precision measuring equipment, serving as a benchmark for evaluating the robot's positioning accuracy.
[0045] Prepare a data recording system for the test, including a high-performance computer, data storage devices, and data analysis software. Also, prepare safety precautions, such as safety ropes and cushioning pads, to prevent accidents involving the robot during testing.
[0046] In step S2, select no fewer than 5 reference points on each floor, instruct the robot to move to each point, record the robot's self-localization coordinates and the actual coordinates measured by the motion capture system, and calculate the positioning error.
[0047] The reference point is a fixed location selected on each floor of the test environment, serving as the evaluation benchmark for the robot's positioning accuracy and navigation capability. The robot's self-positioning coordinates are the current position coordinates calculated by the robot through its own onboard positioning system. The motion capture system's measured coordinates are the coordinates of the robot's actual position captured in real time by a high-precision external positioning device. The positioning error is the deviation between the robot's self-positioning coordinates and the motion capture system's measured coordinates.
[0048] It is understood that in this embodiment of the application, no less than 5 reference points are selected on each floor and the positioning error is calculated. By comparing the multi-area coordinate sampling with the high-precision measured benchmark, the robot's self-positioning deviation is quantified. This can not only comprehensively cover the positioning capability assessment of typical scenarios, but also accurately locate the algorithm's shortcomings through error data, providing a quantitative basis for the optimization of the positioning module and realizing a closed loop of accuracy improvement from testing to application.
[0049] In this embodiment of the application, the formula for calculating the positioning error is:
[0050] ;
[0051] in, Use it as a coordinate system for self-positioning. These are the measured coordinates.
[0052] Specifically, at least five reference points should be selected on each floor. These reference points should be distributed in different functional areas and corridors to evaluate the robot's positioning accuracy in different environments.
[0053] Reflective markers are installed on the robot for tracking by the motion capture system. These markers should be installed in key locations such as the robot's head, torso, and feet to accurately capture the robot's position and orientation.
[0054] The motion capture system is calibrated to ensure measurement accuracy better than 1 millimeter. The calibration process includes camera position calibration, coordinate system calibration, and marker point recognition calibration.
[0055] The robot is instructed to move to each reference point in sequence. After reaching each reference point, the robot's self-localization result (the robot's perceived location coordinates) and the motion capture system's measurement result (the robot's actual location coordinates) are recorded.
[0056] Calculate the Euclidean distance between the robot's self-localized position and its actual position as the localization error. Calculate the average, maximum, and standard deviation of the localization error for each floor to evaluate the robot's localization accuracy across different floors.
[0057] Analyze the differences in positioning accuracy across different floors and functional areas, and assess the impact of floor and environmental factors on positioning accuracy.
[0058] In step S3, the robot is randomly placed on different floors, instructed to identify the current floor, the identification result is recorded, and the floor identification accuracy and average identification time are calculated.
[0059] It is understood that the embodiments of this application conduct floor recognition tests by randomly placing robots to eliminate interference from positional patterns, quantify recognition accuracy and time consumption, which can expose the recognition robustness defects in complex environments and provide data support for algorithm optimization, ensuring that the robot can quickly and accurately complete floor positioning in real-world scenarios.
[0060] Specifically, the robot is placed at random locations on different floors without being informed of its current floor. The robot needs to identify its current floor through an environmental perception and positioning system.
[0061] The robot is instructed to identify its current floor, and the identification result and the actual floor are recorded. The floor identification accuracy is calculated, which is the number of correct identifications divided by the total number of tests.
[0062] To assess the impact of different environmental conditions on floor recognition capabilities, tests should be conducted at different times (day and night), under different lighting conditions (normal lighting, low light, and strong light), and under different pedestrian densities (no one, sparse pedestrian flow, and dense pedestrian flow).
[0063] Record the time required for the robot to identify floors and evaluate its recognition speed. Simultaneously, record the floor recognition strategies used by the robot, such as visual feature recognition, map matching, and height measurement, and analyze the effectiveness of different strategies.
[0064] In step S4, a cross-floor staircase navigation task is set up, requiring the robot to travel from one floor to another via the stairs, and the staircase navigation performance is evaluated.
[0065] The stair navigation performance includes stair recognition success rate, climbing speed, posture stability, and positioning continuity. Positioning continuity is the jump value of positioning error when changing floors.
[0066] It is understood that the embodiments of this application set up a cross-floor staircase navigation task. By simulating a staircase passage scenario in a real-world setting, the robot's path planning, motion control, and environmental adaptability during the process of going up and down stairs can be evaluated. This can expose the robot's deficiencies in stair slope recognition, step height perception, obstacle avoidance strategies, etc. At the same time, by quantifying indicators such as path efficiency and motion stability, a basis can be provided for optimizing the robot's staircase passage algorithm, ensuring that it can safely and efficiently complete cross-floor movement in real-world scenarios.
[0067] Specifically, such as Figure 3 As shown, a cross-floor navigation task is set up, requiring the robot to travel from one floor to another via stairs. The task includes four phases: stair recognition, approach, climbing, and departure.
[0068] Before testing, ensure the stairwell area is well-lit and remove any obstacles that may interfere with the test. Set up additional motion capture cameras in the stairwell area to ensure complete capture of the stair climbing process.
[0069] The robot is instructed to start from the starting point, autonomously find and identify stairs, and then use the stairs to reach the designated location on the target floor. The robot's trajectory, speed, and attitude changes are recorded throughout the process.
[0070] Evaluate the following indicators:
[0071] Staircase recognition success rate: The percentage of times the robot correctly identifies the location and direction of stairs;
[0072] Staircase approach accuracy: The deviation between the robot's position and orientation when approaching the stairs and the ideal approach path;
[0073] Stair climbing speed: The average speed at which the robot climbs stairs (steps / minute);
[0074] Stair climbing stability: The maximum deviation angle of the robot's posture during the climbing process;
[0075] Staircase departure accuracy: The deviation between the robot's position and orientation when leaving the stairs and the ideal departure path;
[0076] Location continuity: The continuity and smoothness of location results during floor transitions.
[0077] Therefore, each floor combination (such as first floor to second floor, second floor to first floor, etc.) was tested 3 times, and the average value and standard deviation of each indicator were calculated.
[0078] In step S5, a cross-floor elevator navigation task is set up, requiring the robot to use the elevator to reach another floor and evaluate the elevator navigation performance.
[0079] Elevator navigation performance includes elevator recognition success rate, call success rate, entry accuracy, and total navigation time.
[0080] It is understood that the embodiments of this application set up a cross-floor elevator navigation task. By simulating the entire interaction process between the robot and the elevator, its environmental perception, device interaction and task planning capabilities are evaluated. This can expose the robot's defects in elevator button recognition, door status judgment and elevator process logic. At the same time, it can quantify indicators such as interaction success rate and elevator efficiency, provide data support for optimizing the robot's elevator interaction algorithm, and ensure that it can autonomously complete the cross-floor elevator task in real-world scenarios.
[0081] Specifically, such as Figure 4As shown, a cross-floor navigation task is set up, requiring the robot to travel from one floor to another via elevator. The task includes stages such as elevator identification, calling, entering, selecting the target floor, waiting, and exiting the elevator.
[0082] Before testing, ensure the elevator is functioning properly and install motion capture cameras inside and outside the elevator to capture the robot's movements. Due to potential GPS signal shielding and magnetic field interference inside the elevator, special attention needs to be paid to the robot's positioning performance in this environment.
[0083] The robot starts from the starting point, autonomously locates and identifies the elevator, calls for the elevator, waits for the elevator to arrive, enters the elevator, selects the target floor (via button operation or voice command), waits for the elevator to reach the target floor, and then exits the elevator to reach the designated location. The robot's movement trajectory, speed, and operational behavior are recorded throughout the entire process.
[0084] Evaluate the following indicators:
[0085] Elevator recognition success rate: The percentage of elevators that the robot correctly identifies.
[0086] Elevator call success rate: The percentage of times a robot successfully calls an elevator;
[0087] Elevator entry accuracy: The deviation between the robot's position and posture when entering the elevator and the elevator's centerline;
[0088] Floor selection success rate: The percentage of times the robot successfully selects the target floor;
[0089] Elevator exit recognition success rate: The percentage of times the robot correctly identifies the exit after the elevator reaches the target floor;
[0090] Elevator departure accuracy: The deviation between the robot's position and orientation when leaving the elevator and the ideal departure path;
[0091] Total elevator navigation time: The total time from the starting point to the destination.
[0092] Therefore, each floor was tested three times, and the average and standard deviation of each indicator were calculated.
[0093] In step S6, a complex multi-floor navigation task is set up, including multiple target points and floor transitions, and the overall navigation capability is evaluated.
[0094] The comprehensive navigation capability includes task completion rate, target point positioning accuracy, path optimization rate, and the rationality of floor switching strategy. The rationality of floor switching strategy is the percentage of times that the elevator / stairs selection meets the needs of the scenario.
[0095] It is understood that the embodiments of this application set up complex navigation tasks involving multiple targets and multiple floors. By simulating real cross-floor and multi-point operation scenarios, the comprehensive capabilities of robot path planning, floor transition and global scheduling are evaluated, and its shortcomings in multi-target decision-making, dynamic obstacle avoidance and other scenarios are exposed. Quantitative indicators such as task efficiency are used to provide a basis for optimizing the global navigation algorithm, so as to ensure efficient handling of cross-floor requirements in practical applications.
[0096] Specifically, such as Figure 5 As shown, a complex multi-floor navigation task is set up, including multiple destinations and floor transitions. For example, "from the first-floor lobby to the second-floor conference room, then to the third-floor offices, and finally back to the first-floor lobby." The task should simulate a real-world application scenario, including navigation between different functional areas and multiple floor transitions.
[0097] Before testing, ensure all equipment in the test environment is functioning properly and remove any temporary obstacles that may affect the test. Set up cameras at key locations to record the robot's navigation process.
[0098] The robot is instructed to complete navigation tasks, and the path planning, execution, and adjustments during the navigation process are recorded. Special attention is paid to the robot's decisions and behaviors at floor transition points (such as stairwells and elevator entrances), as well as the continuity of navigation between different floors.
[0099] Evaluate the following indicators:
[0100] Navigation task completion rate: The percentage of navigation tasks that are successfully completed;
[0101] Navigation accuracy: the positional deviation upon reaching each target point;
[0102] Navigation time: The time required to complete the entire navigation task;
[0103] Path optimization rate: The ratio of the actual path length to the theoretical shortest path length;
[0104] Floor switching strategy: The rationality of the robot's decision to choose between stairs or elevator;
[0105] Navigation continuity: The continuity and smoothness of navigation before and after floor transitions;
[0106] Environmental adaptability: The stability of navigation performance under different environmental conditions (such as changes in lighting and changes in pedestrian density).
[0107] Therefore, each integrated navigation task was tested three times, and the average and standard deviation of each indicator were calculated.
[0108] In step S7, a weighted evaluation model is constructed based on positioning error, floor recognition accuracy, stair navigation performance, elevator navigation performance, and comprehensive navigation capability. A comprehensive score is calculated based on each indicator, and a test report is generated.
[0109] The weighted evaluation model is an assessment system that assigns different weights to indicators such as positioning error, floor recognition accuracy, stair navigation performance, elevator navigation performance, and comprehensive navigation capability, and calculates a comprehensive score through a mathematical model.
[0110] It is understood that the embodiments of this application construct a weighted evaluation model and generate a test report. By integrating multi-dimensional indicators such as positioning error and recognition accuracy and assigning weights to calculate a comprehensive score, fragmented data is transformed into a systematic evaluation, which objectively reflects the quality of robot navigation capabilities and provides precise direction for technological iteration with data charts and optimization suggestions.
[0111] Specifically, such as Figure 6 As shown, the test data processing flow includes four stages: data collection, data preprocessing, indicator calculation, and result analysis.
[0112] During the data collection phase, data from the motion capture system, internal robot sensors, and external observation records are collected into the data processing system.
[0113] In the data preprocessing stage, the raw data is filtered, calibrated, and time-synchronized to eliminate noise and system errors, ensuring the accuracy and consistency of the data.
[0114] In the indicator calculation phase, the performance indicators of each test are calculated according to the preset evaluation indicator system, including positioning accuracy, floor recognition accuracy, stair navigation performance, elevator navigation performance, and overall navigation performance.
[0115] In the results analysis phase, statistical analysis was performed on the calculated indicators, including the mean, standard deviation, and coefficient of variation, to assess the stability and reliability of the robot's performance. Simultaneously, correlation analysis was conducted to explore the relationships between different indicators and the impact of environmental factors on performance.
[0116] In this embodiment of the application, the formula for calculating the comprehensive score based on various indicators is as follows:
[0117] Overall score = Basic positioning performance score * 20% + Floor recognition ability score * 15% + Staircase navigation ability score * 25% + Elevator navigation ability score * 25% + Overall navigation ability score * 15%.
[0118] Specifically, such as Figure 7 As shown, the evaluation index system includes five dimensions: basic positioning performance, floor recognition capability, staircase navigation capability, elevator navigation capability, and comprehensive navigation capability. Each dimension has multiple specific indicators, and the score for that dimension is calculated through a weighted average.
[0119] The overall evaluation score is calculated using the following formula:
[0120] Overall score = Basic positioning performance score × 20% + Floor recognition ability score × 15% + Staircase navigation ability score × 25% + Elevator navigation ability score × 25% + Overall navigation ability score × 15%
[0121] Depending on the needs of different application scenarios, the weights of each dimension can be adjusted to obtain evaluation results that better suit specific application requirements. For example, for service robots that mainly use elevators in office buildings, the weight of elevator navigation ability can be increased; while for shopping guide robots that mainly use stairs and escalators in shopping malls, the weight of stair navigation ability can be increased.
[0122] The final test report includes basic information about the test object, test environment conditions, a description of the test process, test results, data analysis, and improvement suggestions. The test report should objectively and comprehensively reflect the robot's multi-floor navigation capabilities, providing a scientific basis for robot design, optimization, and application.
[0123] In this embodiment of the application, it also includes: comparing and analyzing the test report with human navigation data under the same task, and the comparison dimensions include navigation time, path optimization rate and rationality of floor conversion decision.
[0124] It is understood that this application's embodiments compare and analyze test reports with human navigation data, establishing a reference system between robot navigation capabilities and human cognitive levels through quantitative comparisons of dimensions such as navigation time, path optimization rate, and floor transition decisions. This operation can intuitively expose the robot's shortcomings in dynamic environment adaptation and unstructured decision-making. Simultaneously, through reverse analysis of efficient human navigation strategies, it provides bio-inspired support for optimizing robot path planning algorithms and improving decision-making flexibility, ultimately pushing robot navigation performance closer to human intelligence levels and enhancing the rationality of task execution and environmental adaptability in real-world scenarios.
[0125] This application proposes a method for testing the multi-floor navigation capabilities of humanoid robots. By designing specialized test tasks such as stair climbing and elevator use, and combining quantitative indicators such as floor recognition accuracy and navigation performance, the evaluation dimension is extended from a two-dimensional plane to a three-dimensional space, filling the technical gap in the evaluation of vertical movement and dynamic floor transition capabilities. Simultaneously, by utilizing indicators such as positioning continuity and the rationality of floor transition strategies, a three-dimensional evaluation system covering the entire process of multi-story building navigation is constructed, significantly enhancing the fit between test results and actual application scenarios. Furthermore, by accurately calculating positioning errors and linking them with the target point positioning accuracy indicators in the comprehensive navigation task, the impact of complex environments on robot positioning is analyzed in depth, providing precise data support for algorithm optimization and hardware upgrades. The constructed weighted evaluation model incorporates core performance indicators into the comprehensive score calculation, forming a standardized evaluation paradigm and enabling the scientific quantification and horizontal comparability of the navigation capabilities of different robots in multi-story building environments. Therefore, this solves the problems in existing technologies such as the lack of vertical movement capability evaluation, the inability to evaluate floor recognition and transition capabilities, inaccurate positioning accuracy in multi-story environments, and poor navigation continuity during floor transitions.
[0126] The following will illustrate a specific embodiment of a method for testing the multi-floor navigation capability of a humanoid robot, such as... Figure 8 As shown, the specific content is as follows:
[0127] 1.1 Test Environment Preparation:
[0128] Select a standard test building with at least two floors, including stairs, elevators, and various functional areas; install a high-precision motion capture system on each floor of the test building to record the robot's actual position and posture; set up standardized reference points within the test building to accurately measure and record their three-dimensional coordinates; and prepare the data recording system and safety protection devices required for the test.
[0129] 1.2 Basic Positioning Accuracy Test:
[0130] Select at least 5 reference points on each floor and instruct the robot to move to each reference point in sequence; record the robot's self-localization results and the measurement results of the motion capture system; calculate the deviation between the robot's self-localized position and the actual position, and evaluate the positioning accuracy on each floor.
[0131] 1.3 Floor Recognition Ability Test: Place the robot at random locations on different floors without informing it of its current floor; instruct the robot to recognize its current floor; record the consistency between the robot's floor recognition result and the actual floor; repeat the test 10 times and calculate the floor recognition accuracy.
[0132] 1.4 Staircase Navigation Test: Set up a cross-floor navigation task, requiring the robot to travel from one floor to another via stairs; record the robot's staircase recognition, approach, climbing, and departure processes; measure the speed, stability, and success rate of staircase climbing; evaluate the changes in positioning accuracy during staircase climbing; test each floor combination 3 times.
[0133] 1.5 Elevator Navigation Test: Set up a cross-floor navigation task, requiring the robot to travel from one floor to another via the elevator; record the robot's elevator recognition, calling, entering, selecting the target floor, waiting, and exiting the elevator; measure the success rate and average completion time of elevator use; evaluate the changes in positioning accuracy inside the elevator and at the elevator door; conduct 3 tests for each floor combination.
[0134] 1.6 Comprehensive Navigation Task Test: Set up complex multi-floor navigation tasks, including multiple target points and floor transitions; instruct the robot to complete the navigation task, and record the path planning, execution, and adjustments during the navigation process; measure the completion rate, average completion time, and navigation accuracy of the navigation task; evaluate the robot's navigation continuity during different floors and floor transitions; each task is tested 3 times.
[0135] 1.7 Data Processing and Evaluation: Calculate the average value, standard deviation, and coefficient of variation of each test indicator; comprehensively evaluate the robot's multi-floor navigation capability according to the preset evaluation criteria; generate a test report, including test results, data analysis, and improvement suggestions.
[0136] In summary, the embodiments of this application provide a comprehensive testing method. Through standardized procedures and evaluation indicators, it fills the technical gap in evaluating the multi-story building navigation capabilities of humanoid robots, making the multi-story navigation performance of different robots comparable and providing a scientific basis for design and selection. At the same time, it conducts tests on key capabilities such as floor recognition, stair and elevator navigation, and comprehensive navigation tasks, which can not only comprehensively evaluate the robot's adaptability in multi-story environments, but also improve the practicality of the results through tests that are close to actual application scenarios, thus clarifying the direction for optimizing robot navigation performance.
[0137] Next, referring to the accompanying drawings, a multi-floor navigation capability testing system for a humanoid robot according to an embodiment of this application is described.
[0138] Figure 9 This is a block diagram of a humanoid robot multi-floor navigation capability testing system according to an embodiment of this application.
[0139] like Figure 9As shown, the humanoid robot multi-floor navigation capability testing system 10 includes: an environment configuration module 100, a positioning test module 200, a floor recognition module 300, a stair navigation module 400, an elevator navigation module 500, a comprehensive task module 600, and a data analysis module 700.
[0140] The system includes the following modules: Environment Configuration Module 100 for configuring the test building, deploying the motion capture system, and standardizing reference points; Positioning Test Module 200 for performing basic positioning accuracy tests and calculating positioning errors; Floor Recognition Module 300 for evaluating the robot's floor recognition capabilities and environmental adaptability; Staircase Navigation Module 400 for testing the robot's staircase navigation performance and generating evaluation metrics; Elevator Navigation Module 500 for testing the robot's elevator navigation performance and generating evaluation metrics; Comprehensive Task Module 600 for performing multi-floor comprehensive navigation tasks and evaluating comprehensive performance; and Data Analysis Module 700 for processing test data, calculating comprehensive scores, and generating test reports.
[0141] It should be noted that the foregoing explanation of the embodiment of the humanoid robot multi-floor navigation capability test method also applies to the humanoid robot multi-floor navigation capability test system of this embodiment, and will not be repeated here.
[0142] The humanoid robot multi-floor navigation capability testing system proposed in this application extends the evaluation dimension from a two-dimensional plane to a three-dimensional space by designing specialized test tasks such as stair climbing and elevator use, and combining quantitative indicators such as floor recognition accuracy and navigation performance. This fills the technical gap in evaluating vertical movement and dynamic floor transition capabilities. Simultaneously, by utilizing indicators such as positioning continuity and the rationality of floor transition strategies, a three-dimensional evaluation system covering the entire navigation process in multi-story buildings is constructed, significantly enhancing the fit between test results and actual application scenarios. Furthermore, by accurately calculating positioning errors and linking them with the target point positioning accuracy indicators in comprehensive navigation tasks, the impact of complex environments on robot positioning is analyzed in depth, providing precise data support for algorithm optimization and hardware upgrades. The constructed weighted evaluation model incorporates core performance indicators into the comprehensive score calculation, forming a standardized evaluation paradigm and enabling the scientific quantification and horizontal comparability of navigation capabilities of different robots in multi-story building environments. Therefore, it solves the problems in existing technologies such as the lack of vertical movement capability evaluation, the inability to evaluate floor recognition and transition capabilities, inaccurate positioning accuracy in multi-floor environments, and poor navigation continuity during floor transitions.
[0143] Figure 10 A schematic diagram of the structure of an electronic device provided in an embodiment of this application. The electronic device may include:
[0144] The memory 1001, the processor 1002, and the computer program stored on the memory 1001 and executable on the processor 1002.
[0145] When the processor 1002 executes the program, it implements the method for testing the multi-floor navigation capability of the humanoid robot provided in the above embodiments.
[0146] Furthermore, electronic devices also include:
[0147] Communication interface 1003 is used for communication between memory 1001 and processor 1002.
[0148] The memory 1001 is used to store computer programs that can run on the processor 1002.
[0149] The memory 1001 may include high-speed RAM memory, and may also include non-volatile memory, such as at least one disk storage.
[0150] If the memory 1001, processor 1002, and communication interface 1003 are implemented independently, then the communication interface 1003, memory 1001, and processor 1002 can be interconnected via a bus to complete communication between them. The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized into address buses, data buses, control buses, etc. For ease of representation, Figure 10 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.
[0151] Optionally, in a specific implementation, if the memory 1001, processor 1002, and communication interface 1003 are integrated on a single chip, then the memory 1001, processor 1002, and communication interface 1003 can communicate with each other through an internal interface.
[0152] The processor 1002 may be a central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of this application.
[0153] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.
[0154] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this application, "N" means at least two, such as two, three, etc., unless otherwise explicitly specified.
[0155] Any process or method described in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or N executable instructions for implementing custom logic functions or processes, and the scope of the preferred embodiments of this application includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as should be understood by those skilled in the art to which embodiments of this application pertain.
[0156] It should be understood that the various parts of this application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, the N steps or methods can be implemented using software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or more of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.
[0157] Those skilled in the art will understand that all or part of the steps of the methods described in the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it includes one or a combination of the steps of the method embodiments.
Claims
1. A method for testing the multi-floor navigation capability of a humanoid robot, characterized in that, include: Select the target testing environment; Select no fewer than 5 reference points on each floor, instruct the robot to move to each point, record the robot's self-localization coordinates and the actual coordinates measured by the motion capture system, and calculate the positioning error. The robot is randomly placed on different floors, instructed to identify the current floor, the identification results are recorded, and the floor identification accuracy and average identification time are calculated. Set up a cross-floor staircase navigation task, requiring the robot to travel from one floor to another via stairs, and evaluate the staircase navigation performance. The staircase navigation performance includes staircase recognition success rate, climbing speed, posture stability, and positioning continuity. The posture stability is the maximum deviation of the torso tilt angle during climbing, and the positioning continuity is the jump value of the positioning error when changing floors. Set up a cross-floor elevator navigation task, requiring the robot to use the elevator to reach another floor and evaluate the elevator navigation performance. The elevator navigation performance includes elevator recognition success rate, call success rate, entry accuracy, and total navigation time. Set up complex multi-floor navigation tasks, including multiple target points and floor transitions, and evaluate the comprehensive navigation capabilities. The comprehensive navigation capabilities include task completion rate, target point positioning accuracy, path optimization rate, and the rationality of floor transition strategies. The rationality of floor transition strategies is the percentage of times that the elevator / stairs selection meets the needs of the scenario. A weighted evaluation model is constructed based on the positioning error, the floor recognition accuracy, the stair navigation performance, the elevator navigation performance, and the overall navigation capability. A comprehensive score is calculated based on each indicator, and a test report is generated.
2. The method for testing the multi-floor navigation capability of a humanoid robot according to claim 1, characterized in that, The target testing environment is a standard testing building with at least two floors, with a high-precision motion capture system and standardized reference points set up on each floor.
3. The method for testing the multi-floor navigation capability of a humanoid robot according to claim 2, characterized in that, The high-precision motion capture system includes at least 8 infrared cameras covering the test area on each floor, with a measurement accuracy better than 1mm; there are no fewer than 10 standardized reference points on each floor, distributed in offices, corridors, and stairwells, and their three-dimensional coordinates are measured using a total station.
4. The method for testing the multi-floor navigation capability of a humanoid robot according to claim 1, characterized in that, The formula for calculating the positioning error is: ; in,( () as self-positioning coordinates, ( () represents the measured coordinates.
5. The method for testing the multi-floor navigation capability of a humanoid robot according to claim 1, characterized in that, The formula for calculating the overall score based on each indicator is as follows: Overall score = Basic positioning performance score * 20% + Floor recognition ability score * 15% + Staircase navigation ability score * 25% + Elevator navigation ability score * 25% + Overall navigation ability score * 15%.
6. The method for testing the multi-floor navigation capability of a humanoid robot according to claim 1, characterized in that, Also includes: The test report was compared and analyzed with navigation data of humans performing the same task. The comparison dimensions included navigation time, path optimization rate, and the rationality of floor transfer decisions.
7. A system for testing the multi-floor navigation capability of a humanoid robot as described in claim 1, characterized in that, include: The environment configuration module is used to configure the test building, deploy the motion capture system, and standardize reference points; The positioning test module is used to perform basic positioning accuracy tests and calculate positioning errors; The floor recognition module is used to evaluate the robot's floor recognition capabilities and environmental adaptability. The stair navigation module is used to test the robot's stair navigation performance and generate evaluation metrics. The elevator navigation module is used to test the robot's elevator navigation performance and generate evaluation metrics. The integrated task module is used to perform multi-floor integrated navigation tasks and evaluate overall performance; The data analysis module is used to process test data, calculate the overall score, and generate test reports.
8. An electronic device, characterized in that, include: The system includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the method for testing the multi-floor navigation capability of a humanoid robot as described in any one of claims 1-6.