Dynamic tag tracking method for a perception-limited RFID robot and apparatus therefor

Abstract

The application relates to the technical field of robot tracking, and discloses a dynamic tag tracking method and device for an RFID robot based on perception limitation. The robot is provided with an RFID antenna for detecting a target with an RFID tag; the dynamic tag tracking method comprises the following steps: acquiring phase data for representing the relative position of the RFID tag and the RFID antenna; performing phase unfolding on the phase data, and performing phase correction on the absolute value of the unfolded phase; estimating a virtual candidate position of the RFID tag according to the position information of the RFID tag at the previous moment; determining the position information of the RFID tag at the current moment based on the matching condition of the unfolded phase of the RFID antenna and the phase of the virtual candidate position, so as to obtain the relative position information of the RFID tag and the robot; constructing a potential energy function based on the relative position information of the RFID tag and the robot and the detection constraint of the RFID antenna, and introducing a cost function to calculate the reference tracking speed of the robot. In this way, the RFID robot with the perception limitation can track the target more accurately.

Description

Technical Field

[0001] This application relates to the field of robot tracking technology, and for example to a dynamic tag tracking method and apparatus based on a perception-limited RFID robot. Background Technology

[0002] Mobile robot-based target tracking technology has been successfully applied in various fields, such as healthcare, food service, industry, and others. This has also fueled higher expectations for robot-based tracking technologies, such as improving computational efficiency and protecting data privacy while ensuring basic tracking functionality.

[0003] Currently, most robot tracking systems rely on optical and acoustic sensors, such as cameras and radar. As devices capable of capturing high-resolution images, cameras can acquire rich environmental information. However, the use of cameras also faces many problems, such as limited performance in low light and the risk of privacy exposure. These issues limit the application of camera-based robot tracking in certain scenarios. Unlike cameras, radar or lasers are unaffected by lighting conditions, making them suitable for various lighting scenarios. However, they struggle to distinguish between objects with similar appearances and cannot detect objects with occlusion.

[0004] It should be noted that the information disclosed in the background section above is only used to enhance the understanding of the background of this application, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0005] To provide a basic understanding of some aspects of the disclosed embodiments, a brief summary is given below. This summary is not intended as a general commentary, nor is it intended to identify key / important components or describe the scope of protection of these embodiments, but rather as a prelude to the detailed description that follows.

[0006] This disclosure provides a dynamic tag tracking method and apparatus based on a perception-limited RFID robot, which can more accurately track a specific target in complex environments (such as obstruction, low light, similar appearance, etc.).

[0007] In some embodiments, the dynamic tag tracking method based on a perception-limited RFID robot includes:

[0008] The robot is equipped with an RFID antenna for detecting targets with RFID tags. The dynamic tag tracking method includes: acquiring phase data representing the relative position of the RFID tag and the RFID antenna; performing phase unrolling on the phase data and phase correction on the absolute value of the unrolled phase; estimating a virtual candidate position of the RFID tag based on the position information of the RFID tag at the previous moment; determining the current position information of the RFID tag based on the matching of the unrolled phase of the RFID antenna and the phase of the virtual candidate position, so as to obtain the relative position information of the RFID tag and the robot; constructing a potential energy function based on the relative position information of the RFID tag and the robot and the detection constraints of the RFID antenna, and introducing a cost function to calculate the reference tracking speed of the robot.

[0009] In some embodiments, the dynamic tag tracking device based on a perception-limited RFID robot includes: a processor and a memory storing program instructions, the processor being configured to execute the aforementioned dynamic tag tracking method based on a perception-limited RFID robot when running the program instructions.

[0010] In some embodiments, the robot includes: a robot body; an RFID antenna mounted on the robot body for detecting targets affixed with RFID tags; and a dynamic tag tracking device for a perception-limited RFID robot as described above, mounted on the robot body.

[0011] The dynamic tag tracking method and apparatus based on a perception-limited RFID robot provided in this disclosure can achieve the following technical effects:

[0012] First, phase data representing the relative position of the RFID tag and RFID antenna is acquired. Then, phase unfolding and phase correction are performed on the phase data to obtain more accurate phase data, which is beneficial for more precise determination of the location of a specific RFID tag. Based on the RFID tag's previous position, virtual candidate positions of the RFID tag are estimated, thus determining the RFID tag's current position information. Considering the RFID antenna detection constraints and the robot's own collision avoidance requirements, a potential energy function is constructed based on the relative position information of the RFID tag and robot and the RFID antenna detection constraints, and then the robot's reference tracking speed is calculated. In this way, the perception-limited RFID robot can adaptively control its tracking speed to achieve both specific target tracking and adaptive collision avoidance, enabling it to operate in scenarios with occlusion, low light, and no light, while also protecting user privacy.

[0013] The above general description and the description below are exemplary and illustrative only and are not intended to limit this application. Attached Figure Description

[0014] One or more embodiments are illustrated by way of example with reference to the accompanying drawings. These illustrations and drawings do not constitute a limitation on the embodiments. Elements having the same reference numerals in the drawings are shown as similar elements. The drawings are not to be scaled. And wherein:

[0015] Figure 1 This is a schematic diagram of the RFID-based collaborative collision avoidance and tracking system provided in this embodiment;

[0016] Figure 2 This is a schematic diagram of the first dynamic tag tracking method based on a perception-limited RFID robot provided in the embodiments of this disclosure;

[0017] Figure 3 This is a schematic diagram of a method for phase unfolding phase data provided in an embodiment of this disclosure;

[0018] Figure 4 This is a schematic diagram of a method for estimating the virtual candidate location of an RFID tag based on the location information of the RFID tag at the previous moment, provided in an embodiment of this disclosure.

[0019] Figure 5 This is a schematic diagram of a method for determining the current location information of an RFID tag based on the matching of the deployed phase of the RFID antenna and the phase of the virtual candidate location, provided in an embodiment of this disclosure.

[0020] Figure 6 This is a schematic diagram of a method for calculating the robot's reference tracking speed based on the relative position information of the RFID tag and the robot and the detection constraints of the RFID antenna, provided in this embodiment of the disclosure;

[0021] Figure 7 This is a schematic diagram of the second dynamic tag tracking method based on a perception-limited RFID robot provided in the embodiments of this disclosure;

[0022] Figure 8 This is a schematic diagram illustrating the positioning performance evaluation of different trajectories in normal scenes, dark scenes, and occluded scenes provided in the embodiments of this disclosure;

[0023] Figure 9 It is the cumulative distribution function (CDF) of the x-axis and y-axis positioning errors in different scenarios provided in the embodiments of this disclosure.

[0024] Figure 10 This is a scatter plot of the matching function and the corresponding estimated label position provided in the embodiments of this disclosure;

[0025] Figure 11 This is a graph of distance and azimuth provided in an embodiment of this disclosure;

[0026] Figure 12 This is the first dynamic tag tracking device based on a perception-limited RFID robot provided in this disclosure embodiment;

[0027] Figure 13 This is the second dynamic tag tracking device based on a perception-limited RFID robot provided in this disclosure embodiment.

[0028] Figure label:

[0029] 10. Robot body; 20. RFID antenna;

[0030] 120. First dynamic tag tracking device based on a perception-limited RFID robot; 121. Phase data acquisition module; 122. Phase data processing module; 123. Tag position estimation module; 124. Tag position determination module; 125. Tracking speed calculation module;

[0031] 130. Second dynamic tag tracking device based on perception-limited RFID robot; 131. Processor; 132. Memory; 133. Communication interface; 134. Bus. Detailed Implementation

[0032] To provide a more detailed understanding of the features and technical content of the embodiments of this disclosure, the implementation of the embodiments of this disclosure will be described in detail below with reference to the accompanying drawings. The accompanying drawings are for illustrative purposes only and are not intended to limit the embodiments of this disclosure. In the following technical description, for ease of explanation, several details are used to provide a full understanding of the disclosed embodiments. However, one or more embodiments may still be implemented without these details. In other cases, well-known structures and devices may be simplified in their depiction to simplify the drawings.

[0033] The terms "first," "second," etc., used in the specification, claims, and accompanying drawings of this disclosure are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate for the embodiments of this disclosure described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion.

[0034] Unless otherwise stated, the term "multiple" means two or more.

[0035] In this embodiment of the disclosure, the character " / " indicates that the objects before and after it are in an "or" relationship. For example, A / B means: A or B.

[0036] The term "and / or" describes an association between objects, indicating that three relationships can exist. For example, A and / or B means: A or B, or A and B.

[0037] The term "correspondence" can refer to an association or binding relationship. The correspondence between A and B means that there is an association or binding relationship between A and B.

[0038] This disclosure provides an embodiment of a robot, including a robot body 10. The robot body 10 is equipped with multiple RFID (Radio Frequency Identification) antennas 20 and an RFID reader. Specifically, in conjunction with... Figure 1 As shown, four horizontally arranged RFID antennas 20 are mounted on the robot body 10 to detect targets (such as people) with RFID tags attached. After the RFID antennas transmit radio frequency signals, the RFID tags receiving the radio frequency signals are activated from their dormant state. Then, the RFID tags use the radio frequency antennas to send the stored phase data to the RFID reader. The main purpose is to enable the robot to track specific targets using RFID, thereby ensuring that the target remains within the detection constraints of the RFID antennas.

[0039] Alternatively, the robot can be a wheeled robot, a legged robot, a tracked robot, or a wheel-legged hybrid robot.

[0040] A. Constraints of cooperative obstacle avoidance tracking

[0041] According to the Friis transmission equation, the transmit and receive power of an antenna are related to the gain of the transmitting antenna, the gain of the receiving antenna, the signal wavelength, and the distance between the transmitting and receiving antennas. The relationship between transmit and receive power can be expressed as:

[0042] (1)

[0043] in, P r This indicates the antenna's received power. P t This indicates the antenna's transmit power. G t This indicates the gain of the transmitting antenna. G r This indicates the receiver gain. Indicates the wavelength of the signal. d This indicates the distance between the transmitting antenna and the receiver.

[0044] In a Cartesian coordinate system, the position of the RFID antenna is represented as (x a , y a , z a The robot's position is represented as ( x r , y r , z r The location of the target (the tracked object) is represented as ( x o , y o , z o ) indicates that the location of the RFID tag is defined as ( x tag , y tag , z tag Assume the RFID antenna and RFID tag are at the same height (i.e., z a = z tag This can be viewed as the projection of the 3D coordinates of the antenna and the target onto a 2D plane. Furthermore, by projecting the position of the RFID antenna onto a 2D plane, the robot's 2D coordinates can be approximately derived, and similarly, by projecting the position of the RFID tag onto the same plane, the target's 2D coordinates can be derived. In other words, x a = x r , y a = y r , x tag = x o , y tag = y o The distance between the robot and the target. d and azimuth The coordinates between the two in the Cartesian coordinate system can be described as follows:

[0045] (2)

[0046] To ensure the target remains within detection range, the range and azimuth are subject to the following constraints:

[0047] (3)

[0048] in, Indicates the safe distance for collision avoidance. This indicates the maximum effective detection range of the RFID antenna. This indicates the minimum detection azimuth angle of the RFID antenna. This indicates the maximum detection azimuth angle of the RFID antenna.

[0049] B. Kinematic and Dynamic Framework

[0050] In the kinematic equations of a wheeled robot, its velocity along the x / y axis can be described as the components of the velocity along the x / y axis:

[0051] (4)

[0052] in, v This represents the robot's linear velocity.

[0053] The dynamic model of a wheeled robot can be described as: the total torque required to drive a robot to move with a specific acceleration ( This is equivalent to overcoming its inertial force. Coriolis force and centrifugal force ,gravity The sum of these three is:

[0054] (5)

[0055] in, The inertia matrix of the robot. Represents the centripetal force and Coriolis force matrix. Vectors representing gravity and torque. This represents the control input.

[0056] Combination Figure 2 As shown, this disclosure provides a dynamic tag tracking method based on a perception-limited RFID robot, including:

[0057] S101, acquire phase data to characterize the relative positions of the RFID tag and the RFID antenna;

[0058] S102, perform phase expansion on the phase data and perform phase correction on the absolute value of the expanded phase;

[0059] S103, Estimate the virtual candidate location of the RFID tag based on the location information of the RFID tag at the previous moment;

[0060] S104, Based on the matching of the deployed phase of the RFID antenna and the phase of the virtual candidate position, determine the current position information of the RFID tag to obtain the relative position information between the RFID tag and the robot.

[0061] S105: Based on the relative position information of the RFID tag and the robot and the detection constraints of the RFID antenna, a potential energy function is constructed, and a cost function is introduced to calculate the robot's reference tracking speed.

[0062] The robot's RFID reader acquires phase data transmitted by RFID tags, which characterizes the relative position of the RFID tag and RFID antenna. This phase data is then processed. Since the phase value is used to calculate the target's position information, its accuracy needs to be guaranteed. Therefore, phase ambiguity needs to be removed to accurately construct the phase value. First, phase unwrapping is performed, using the original phase at the current moment, the original phase at the last moment, and the unwrapped phase at the last moment to represent the unwrapped phase at the current moment. After phase unwrapping, a continuously changing phase value is obtained. Because RFID is primarily used for identification rather than positioning, the absolute accuracy of its phase information is not high; therefore, phase correction is necessary. Thus, phase correction is performed on the unwrapped phase to obtain the relative position of the RFID tag and RFID antenna, and then the accurate phase value is calculated.

[0063] The basic concept of RFID positioning originates from particle filtering. It determines the RFID tag's position by estimating the matching between the phase information received by the RFID antenna and the phase information estimated based on the RFID tag's position. Virtual candidate positions of the RFID tag are estimated based on its previous position information. Here, there are multiple virtual candidate positions, representing the possible locations the RFID tag might move to at the current moment. The deployed phase of the RFID antenna is matched with the phase of each virtual candidate position, and the current position information of the RFID tag is determined based on the matching results. Combined with the known current position information of the robot, the relative position information of the RFID tag and the robot can be obtained.

[0064] Considering the RFID antenna detection constraints, namely distance and azimuth constraints, a potential energy function is constructed based on the relative position information of the RFID tag and the robot obtained from the aforementioned matching, as well as the RFID antenna detection constraints. The cost function is then introduced into the potential energy function to calculate the robot's reference tracking speed. In this way, the robot's controller controls the locomotion mechanism (such as wheels) to move at the reference tracking speed, thus achieving target tracking.

[0065] The dynamic tag tracking method based on a perception-limited RFID robot provided in this disclosure first acquires phase data representing the relative position of the RFID tag and RFID antenna. Then, the phase data undergoes phase unfolding and phase correction to obtain more accurate phase data, which facilitates more precise determination of the RFID tag's position. Based on the RFID tag's previous position, a virtual candidate position is estimated, thereby determining the RFID tag's current position information. Considering the RFID antenna detection constraints and the robot's collision avoidance requirements, a potential energy function is constructed based on the relative position information of the RFID tag and the robot, and the azimuth and distance constraints between the RFID tag and the mobile robot, to calculate the robot's reference tracking speed. This allows the perception-limited RFID robot to adaptively control its tracking speed while simultaneously achieving specific target tracking and adaptive collision avoidance. It can operate in scenarios with obstructions, low light, or no light, while also protecting user privacy.

[0066] Optionally, combined Figure 3 As shown, S102, phase expansion is performed on the phase data, including:

[0067] S112, calculate the first difference between the original phase at the current time and the original phase at the last time, the second difference between the original phase at the current time minus 2π and the original phase at the last time, and the third difference between the original phase at the current time plus 2π and the original phase at the last time.

[0068] S122, the expanded phase of the phase data at the current moment is represented as the sum of the minimum value among the first difference, the second difference, and the third difference and the expanded phase at the last moment.

[0069] (6)

[0070] in, This represents the original phase at the current moment. Indicates the original phase at the last moment. Indicates the unfolding phase at the final moment. This indicates the unfolded phase at the current moment.

[0071] Optionally, S102, phase correction is performed on the phase data, including:

[0072] When the initial location information of the RFID tag is known, phase correction is performed on the phase data using a static correction method; or,

[0073] When the initial location information of an RFID tag is unknown, a dynamic correction method is used to correct the phase data. The dynamic correction method includes obtaining the position of a stationary RFID tag and then correcting it using a static correction method.

[0074] For static calibration, it is assumed that the initial position of a stationary RFID tag is at a known location. The initial position of RFID antenna i is at a known location point. .

[0075] So, the distance between the RFID antenna and the RFID tag It can be calculated using the coordinates between the two, as shown in the following formula:

[0076] (7)

[0077] Where, ∥... ∥ represents... The L2 norm.

[0078] Based on the mapping relationship between distance and phase, the corresponding calibration phase is related to the distance between the RFID antenna and the RFID tag and the signal wavelength:

[0079] (8)

[0080] in, This represents the wavelength of the signal. Static correction methods are suitable for situations where the initial position information is known. If the initial position information of the RFID tag is unknown, dynamic correction methods can be used to correct the phase data. Dynamic correction methods first obtain the position of a stationary RFID tag, and then use static correction methods for correction.

[0081] Optionally, combined Figure 4 As shown, in step S103, based on the location information of the RFID tag at the previous moment, the virtual candidate location of the RFID tag is estimated, including:

[0082] S113 defines N arbitrary particles within the range (1-,1);

[0083] S123, calculate the product of any particle and the maximum possible distance traveled, and obtain the product;

[0084] S133, the sum of the product and the position of the RFID tag at the previous moment is used to determine the virtual candidate position.

[0085] When an RFID tag changes, its location information also changes. Define N arbitrary particles within the range (1-1):

[0086] (9)

[0087] The maximum possible travel distance of an RFID tag is Therefore, the virtual candidate position is equal to the RFID tag's position at the previous moment plus the product of any particle and the maximum possible movement distance, expressed as:

[0088] (10)

[0089] In this way, multiple virtual candidate locations of the RFID tag can be obtained. The phase of these virtual candidate locations is matched with the deployed phase of the RFID antenna to obtain the optimal virtual candidate location, thereby determining the current location information of the RFID tag.

[0090] Optionally, combined Figure 5 As shown, in step S104, based on the matching of the deployed phase of the RFID antenna and the phase of the virtual candidate location, the current location information of the RFID tag is determined, including:

[0091] S114, Based on the expanded phases of multiple RFID antennas, construct the first normalized vector sequence of phase differences from different RFID antennas expanded in the complex frequency domain;

[0092] S124, Based on the virtual candidate locations, construct a second normalized vector sequence in the complex frequency domain to expand the phase difference from different RFID antennas;

[0093] S134, Based on the matching function related to the normalized spatial domain, calculate the similarity value of the first normalized vector sequence and the second normalized vector sequence in the multidimensional space;

[0094] S144, Determine the current location information of the RFID tag based on the similarity value.

[0095] The aforementioned virtual candidate positions Substituting into formula (7) and combining it with formula (8), we obtain its normalized vector sequence:

[0096] (11)

[0097] in, and The estimated unfolded phases of two different RFID antennas in the phase data from the same RFID tag can be obtained by formula (8).

[0098] Based on the expanded phase of the RFID antenna, a second normalized vector sequence is constructed, the value of which is equal to the phase difference from different RFID antennas expanded in the complex frequency domain.

[0099] (12)

[0100] in, and These represent the unfolded phases of two different RFID antennas in the phase data from the same RFID tag.

[0101] Optionally, the two different RFID antennas mentioned above are the two outermost RFID antennas, i.e. Figure 1 The first and fourth RFID antennas from left to right are shown in the middle.

[0102] Based on the matching function of normalized spatial domain cross-correlation, the similarity values ​​of the first normalized vector sequence and the second normalized vector sequence in multidimensional space are calculated:

[0103] (13)

[0104] Based on similarity value Determine the current location of the RFID tag. This is understandable. The larger the value, the closer the corresponding virtual candidate location is to the current location information of the RFID tag.

[0105] Optionally, based on similarity values, the current location information of the RFID tag is determined, including:

[0106] The information of the virtual candidate location corresponding to the highest similarity value is used as the current location information of the RFID tag. In this way, the most likely current location of the RFID tag can be obtained.

[0107] Optionally, combined Figure 6 As shown in step S105, a potential energy function is constructed based on the relative position information between the RFID tag and the robot, as well as the detection constraints of the RFID antenna. A cost function is then introduced to calculate the robot's reference tracking speed, including:

[0108] S115, Based on the distance and azimuth information between the RFID tag and the robot, construct a tracking error that satisfies the detection constraints;

[0109] S125, construct the potential energy function based on the tracking error, the current position information of the RFID tag, and the detection constraints of the RFID antenna;

[0110] S135, Define the cost function, introduce the cost function into the potential energy function, and take the partial derivative of the potential energy function with respect to the task error to obtain the task error expression;

[0111] S145, Calculate the reference tracking speed based on the task error expression.

[0112] First, a tracking error satisfying detection constraints is constructed based on the distance and azimuth information between the RFID tag and the robot. This distance and azimuth information includes both actual and expected values. Assumptions are made regarding variables. and The expected values ​​are respectively and Define tracking error .

[0113] Calculate the error between the expected value and the actual value:

[0114] (14)

[0115] According to formula (3), formula (14) can be simplified to:

[0116] (15)

[0117] Therefore, by introducing a decay function, equation (15) can be rewritten as:

[0118] (16)

[0119] in,

[0120] (17)

[0121] , . , , , These are the corresponding parameters that satisfy the detection constraints.

[0122] A potential energy function is constructed based on the tracking error and the current position information of the RFID tag. A cost function is defined by combining distance constraints (maximum and minimum distance values) and azimuth constraints (maximum and minimum azimuth values). The definition is as follows:

[0123] (18)

[0124] in, Represents task variables, .

[0125] Then, by introducing a cost function into each corresponding potential energy function, the potential energy function becomes:

[0126] (19)

[0127] in, .

[0128] Find the equation (19) with respect to... The partial derivatives are obtained.

[0129] (20)

[0130] because Since it is continuous, formula (20) can be simplified to:

[0131] (twenty one)

[0132] Define the task error as According to formula (21), the expression for the task error can be expressed as:

[0133] (twenty two)

[0134] according to Definition, reference speed for:

[0135] (twenty three)

[0136] in, This is the proportionality coefficient. .

[0137] Combination Figure 7 As shown, this disclosure provides another dynamic tag tracking method based on a perception-limited RFID robot, including:

[0138] S101, acquire phase data to characterize the relative positions of the RFID tag and the RFID antenna;

[0139] S106 converts the phase angle of the phase data into the corresponding sine and cosine values;

[0140] S107, perform third-order difference calculations on the sine and cosine values ​​respectively;

[0141] S108, reconstruct phase data based on the interpolated sine and cosine values;

[0142] S109, perform phase expansion on the reconstructed phase data and perform phase correction on the absolute value of the expanded phase;

[0143] S103, Estimate the virtual candidate location of the RFID tag based on the location information of the RFID tag at the previous moment;

[0144] S104, Based on the matching of the deployed phase of the RFID antenna and the phase of the virtual candidate position, determine the current position information of the RFID tag to obtain the relative position information between the RFID tag and the robot.

[0145] S105: Based on the relative position information of the RFID tag and the robot and the detection constraints of the RFID antenna, a potential energy function is constructed, and a cost function is introduced to calculate the robot's reference tracking speed.

[0146] To address the issue of insufficient phase data sampling, phase data interpolation is performed. Since phase data is a data set containing... For periodic signals (where the value of N is unknown), direct interpolation does not yield satisfactory results. Therefore, the phase angles contained in the phase data are first converted into corresponding sine and cosine values. Then, third-order interpolation is performed on these sine and cosine values ​​respectively. Finally, the phase data is reconstructed based on the interpolated sine and cosine values. Phase expansion and phase correction are performed on the reconstructed phase data, followed by steps S103 to S105. The phase angles obtained by interpolation in this way are closer to the true values.

[0147] To evaluate the effectiveness of the dynamic tag tracking method based on a perception-limited RFID robot provided in the embodiments of this disclosure, the following evaluation tests were conducted using a low-cost self-assembly robot.

[0148] A. Performance evaluation of radio frequency identification positioning

[0149] To evaluate the performance of the RFID positioning method provided in this embodiment, tests were conducted in different experimental scenarios, including normal, dark, and occluded scenes, to assess the positioning effectiveness of the proposed algorithm. During the tests, the robot was instructed to remain stationary while the position of the RFID tag was calculated. Since the robot is a self-assembling robot, its positioning accuracy cannot match that of industrial robots. To avoid the influence of the robot's own positioning on the RFID positioning error, the robot remained stationary during the positioning performance evaluation, which facilitates comparison with other similar methods. Real-world data was acquired during the tests using a sliding rail device.

[0150] The RFID tag was initially positioned at point (0.5, 1), and the experimental area was 1 m × 0.7 m. Clearly, the proposed positioning algorithm effectively calculated the position of the moving tag. Specifically, in normal scenarios, the average errors along the x-axis and y-axis were 0.031 m and 0.008 m, respectively. In addition to normal scenarios, tests were also conducted in dark and occluded scenarios to investigate the impact of lighting conditions and obstacles on positioning accuracy. Figure 8 The experimental results show that in dark scenes, the average error along the x-axis is 0.028 meters, and the average error along the y-axis is 0.013 meters. In occluded scenes, the average error along the x-axis is 0.023 meters, and the average error along the y-axis is 0.029 meters.

[0151] To more intuitively demonstrate the positioning accuracy of the proposed method in different scenarios, Figure 9 The corresponding cumulative distribution function is provided. This method is compared with similar methods (Table 1), focusing on the following aspects: the object being located, the features used in the method, the use of reference tags, the pre-training requirements, the experimental scenario, and the average error. Compared with other algorithms, the ELM-KF method exhibits higher positioning accuracy (0.059 m). However, this algorithm requires pre-training. While RFID-based positioning methods (0.054 m) and SAR methods (0.06 m (x), 0.08 m (y)) achieve relatively high accuracy, they rely on reference tags for positioning assistance. SILoc and phase decoupling methods have high positioning accuracy, but this method achieves even higher accuracy. Thus, it can be concluded that the positioning method proposed in this embodiment improves accuracy by 30.4% compared to SILoc and by 17.5% compared to the SAR method. See also... Figure 10 The image shows a scatter plot of the matching function and the corresponding tag location estimate. The left image depicts a trajectory, while the right image shows a scatter plot with four points labeled A, B, C, and D. The points within the red boxes in regions A, B, C, and D match the true values ​​better, indicating that highly matched regions can reflect the object's movement. The positioning method provided in this embodiment can effectively estimate changes in RFID tag movement.

[0152] Table 1. Performance comparison between existing positioning methods and the RFID tag positioning method provided in this embodiment.

[0153]

[0154] It should be noted that the method in Table 1 above, EKF-KF, can be found in the literature: Fusion of RSS and Phase Shift using the Kalman Filter for RFID Tracking.

[0155] For RF track information, please refer to the paper Rf-track: Real-time tracking of RFID tags with stationary antennas.

[0156] The RFID-based method can be found in the literature An rfid-based mobile robot localization method combining phase difference and readability.

[0157] For information on the SAR method, please refer to the paper "Robot-based indoor positioning of uhf-rfidtags: The SAR method with multiple trajectories".

[0158] For RF-SLAM, please refer to the paper Rf-SLAM: Uhf-rfid based simultaneous tags mapping and robot localization algorithm for smart warehouse position service.

[0159] For more information on Guide and Tracking, please refer to the literature Guideandtrack: A novel guidance and localization method by a handheld UHF-RFID reader.

[0160] For more information on SILOC, please refer to the literature: Siloc: A speed inconsistency-immune approach to mobile RFID robot localization.

[0161] Phase Decoupling method can be found in the literature Modeling for phase decoupling to detect the orientation and position of moving objects with simple rfid array.

[0162] The proposed method is the method proposed in this invention.

[0163] B. Performance evaluation of tracking methods

[0164] Multiple experimental scenarios: To further evaluate the performance of the robot dynamic label-based tracking method, factors such as similarity, occlusion, and multiple labels were considered. Therefore, two labels were used in this test, one as the target and one as a distractor. To introduce occlusion, the labels were attached to similar-looking foam boxes, making them indistinguishable by visual inspection. In this test, a camera was used to record the trajectories of the target and the robot. The camera's sampling frequency was 30Hz. The value is set to 0.1. It is 1 meter, and It is 3 meters. and The values ​​are respectively 15° and 15°. Expected distance values. The value is 1.5 meters, and the desired directional value is... =0, It is 0.2.

[0165] Figure 11 The graph displays the distance and orientation of the task variables. Taking distance as an example, as the target moves further away, the distance between the robot and the target increases, represented as... Increase. At this point, the robot begins tracking. The distance between the robot and the target begins to decrease, eventually converging to the required distance of 1.5 meters. Similarly, direction... The changing trend is also related to the task variables Same. Overall, the evaluation validated the performance of the adaptive controller with RFID limit awareness.

[0166] Combination Figure 12 As shown, this disclosure provides a first dynamic tag tracking device 120 based on a perception-limited RFID robot, including: a phase data acquisition module 121, a phase data processing module 122, a tag position estimation module 123, a tag position determination module 124, and a tracking speed calculation module 125. The phase data acquisition module 121 is configured to acquire phase data characterizing the relative position of the RFID tag and the RFID antenna. The phase data processing module 122 is configured to perform phase unwrapping on the phase data and perform phase correction on the absolute value of the unwrapped phase. The tag position estimation module 123 is configured to estimate a virtual candidate position of the RFID tag based on the position information of the RFID tag at the previous moment. The tag position determination module 124 is configured to determine the current position information of the RFID tag based on the matching of the unwrapped phase of the RFID antenna and the phase of the virtual candidate position, thereby obtaining the relative position information of the RFID tag and the robot. The tracking speed calculation module 125 is configured to construct a potential energy function based on the relative position information of the RFID tag and the robot and the detection constraints of the RFID antenna, and introduce a cost function to calculate the reference tracking speed of the robot.

[0167] The first dynamic tag tracking device 120 based on a perception-limited RFID robot provided in this disclosure first acquires phase data representing the relative position of the RFID tag and the RFID antenna. Then, it performs phase unfolding and phase correction on the phase data to obtain more accurate phase data, which is beneficial for more precise determination of the RFID tag's position. Based on the RFID tag's previous position, a virtual candidate position of the RFID tag is estimated, thereby determining the RFID tag's current position information. Considering the RFID antenna detection constraints and the robot's own collision avoidance requirements, a potential energy function is constructed based on the relative position information of the RFID tag and the robot, and the azimuth and distance constraints between the RFID tag and the mobile robot, and then the robot's reference tracking speed is calculated. In this way, the perception-limited RFID robot can adaptively control its tracking speed to simultaneously achieve specific target tracking and adaptive collision avoidance, enabling it to operate in scenarios with obstructions, low light, or no light, while also protecting user privacy.

[0168] Combination Figure 13 As shown, this disclosure provides a second dynamic tag tracking device 130 based on a perception-limited RFID robot, including a processor 131 and a memory 132. Optionally, the device 130 may further include a communication interface 133 and a bus 134. The processor 131, communication interface 133, and memory 132 can communicate with each other via the bus 134. The communication interface 133 can be used for information transmission. The processor 131 can call logical instructions in the memory 132 to execute the dynamic tag tracking method based on a perception-limited RFID robot described in the above embodiment.

[0169] Furthermore, the logic instructions in the aforementioned memory 132 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium.

[0170] The memory 132, as a computer-readable storage medium, can be used to store software programs and computer-executable programs, such as program instructions / modules corresponding to the methods in the embodiments of this disclosure. The processor 131 executes functional applications and data processing by running the program instructions / modules stored in the memory 132, thereby realizing the dynamic tag tracking method based on a perception-limited RFID robot in the above embodiments.

[0171] The memory 132 may include a program storage area and a data storage area. The program storage area may store the operating system and applications required for at least one function; the data storage area may store data created based on the use of the terminal device. Furthermore, the memory 132 may include high-speed random access memory and may also include non-volatile memory.

[0172] This disclosure provides a robot, including: a robot body, and the aforementioned first dynamic tag tracking device 120 or second dynamic tag tracking device 130 based on a perception-limited RFID robot. The first dynamic tag tracking device 120 or the second dynamic tag tracking device 130 based on a perception-limited RFID robot is mounted on the robot body. The mounting relationship described herein is not limited to placement inside the robot body, but also includes mounting connections with other components of the robot, including but not limited to physical connections, electrical connections, or signal transmission connections. Those skilled in the art will understand that the first dynamic tag tracking device 120 or the second dynamic tag tracking device 130 based on a perception-limited RFID robot can be adapted to feasible product bodies to achieve other feasible embodiments.

[0173] This disclosure provides a computer-readable storage medium storing computer-executable instructions configured to execute the aforementioned dynamic tag tracking method based on a perception-limited RFID robot.

[0174] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to embodiments of this disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. In some alternative implementations, the functions marked in the blocks may occur in a different order than that shown in the drawings. In the descriptions corresponding to the flowcharts and block diagrams in the accompanying drawings, the operations or steps corresponding to different blocks may also occur in a different order than disclosed in the description; sometimes there is no specific order between different operations or steps.

Claims

1. A dynamic tag tracking method based on a perception-limited RFID robot, characterized in that, The robot is equipped with an RFID antenna for detecting targets with RFID tags; the dynamic tag tracking method includes: S1, Phase data acquisition step: Acquire phase data to characterize the relative position of the RFID tag and the RFID antenna; S2, Phase data processing steps: Phase expansion is performed on the phase data, and phase correction is applied to the absolute value of the expanded phase; the phase expansion of the phase data includes: calculating a first difference between the original phase at the current time and the original phase at the last time, a second difference between the original phase at the current time minus 2π and the original phase at the last time, and a third difference between the original phase at the current time plus 2π and the original phase at the last time; the expanded phase of the phase data at the current time is represented as the sum of the minimum value among the first difference, the second difference, and the third difference, and the expanded phase at the last time. S3, Tag location estimation step: Estimate the virtual candidate location of the RFID tag based on the location information of the RFID tag at the previous moment; S4, Tag Position Determination Step: Based on the matching of the expanded phase of the RFID antenna and the phase of the virtual candidate position, determine the current position information of the RFID tag to obtain the relative position information of the RFID tag and the robot; there are multiple RFID antennas; determining the current position information of the RFID tag based on the matching of the expanded phase of the RFID antenna and the phase of the virtual candidate position includes: constructing a first normalized vector sequence based on the expanded phases of the multiple RFID antennas, expanding the phase difference from different RFID antennas in the complex frequency domain; constructing a second normalized vector sequence based on the virtual candidate position, expanding the phase difference from different RFID antennas in the complex frequency domain; calculating the similarity value of the first normalized vector sequence and the second normalized vector sequence in multidimensional space based on the normalized spatial domain correlation matching function; determining the current position information of the RFID tag based on the similarity value; S5, Tracking speed calculation steps: Based on the relative position information of the RFID tag and the robot and the detection constraints of the RFID antenna, a potential energy function is constructed, and a cost function is introduced to calculate the reference tracking speed of the robot.

2. The dynamic tag tracking method based on a perception-limited RFID robot according to claim 1, characterized in that, Phase correction of the phase data includes: When the initial location information of the RFID tag is known, the phase data is phase-corrected using a static correction method; or, When the initial position information of the RFID tag is unknown, a dynamic correction method is used to perform phase correction on the phase data. The dynamic correction method includes obtaining the position of the stationary RFID tag and then using the static correction method for correction.

3. The dynamic tag tracking method based on a perception-limited RFID robot according to claim 1, characterized in that, The step of estimating the virtual candidate location of the RFID tag based on its previous location information includes: Define N arbitrary particles within the range (-1, 1); Calculate the product of any particle and its maximum possible distance of movement to obtain the product; The product and the sum of the previous location information of the RFID tag are used to determine the virtual candidate location.

4. The dynamic tag tracking method based on a perception-limited RFID robot according to claim 1, characterized in that, Determining the current location information of the RFID tag based on the similarity value includes: The information of the virtual candidate location corresponding to the highest similarity value is used as the current location information of the RFID tag.

5. The dynamic tag tracking method based on a perception-limited RFID robot according to claim 1, characterized in that, The process of constructing a potential energy function based on the relative position information between the RFID tag and the robot, and the detection constraints of the RFID antenna, and introducing a cost function to calculate the robot's reference tracking speed includes: Based on the distance and azimuth information between the RFID tag and the robot, a tracking error that satisfies the detection constraints is constructed; The potential energy function is constructed based on the tracking error, the current position information of the RFID tag, and the detection constraints of the RFID antenna. Define a cost function, introduce the cost function into the potential energy function, and take the partial derivative of the potential energy function with respect to the task error to obtain the task error expression; The reference tracking speed is calculated based on the task error expression.

6. The dynamic tag tracking method based on a perception-limited RFID robot according to any one of claims 1 to 5, characterized in that, After acquiring phase data characterizing the relative positions of the RFID tag and the RFID antenna, and before performing phase unwrapping and phase correction on the phase data, the method further includes: The phase angle of the phase data is converted into corresponding sine and cosine values; The third-order difference between the sine and cosine values ​​is calculated respectively. The phase data is reconstructed based on the interpolated sine and cosine values.

7. A dynamic tag tracking device based on a perception-limited RFID robot, comprising a processor and a memory storing program instructions, characterized in that, The processor is configured to execute, when running the program instructions, the dynamic tag tracking method based on a perception-limited RFID robot as described in any one of claims 1 to 6.

8. A robot, characterized in that, include: The robot itself; An RFID antenna, installed on the robot body, is used to detect targets with RFID tags attached. The dynamic tag tracking device based on a perception-limited RFID robot as described in claim 7 is installed on the robot body.

Images