Positioning method and system, vehicle and intelligent entering and starting system

Abstract

The invention provides a positioning method and system, a vehicle and an intelligent entering and starting system, and relates to the technical field of positioning algorithms, and the positioning method comprises the steps that three anchor points obtain positioning signals sent by an unknown node, respectively calculate corresponding RSSI values, and respectively calculate the distances between the three anchor points and the unknown node according to the corresponding RSSI values; drawing a circle by taking the three anchor points as circle centers and taking the distance between the three anchor points and the unknown node as a radius to obtain three circles, and judging the current state of the three circles according to the position relationship of the three circles; and according to the current three-circle state, respectively carrying out weighted compensation on the three-circle state to obtain the position of the unknown node. The positioning method provided by the invention is improved on the basis of a traditional trilateral positioning algorithm, and corresponding compensation is performed according to different three-circle position states, so that positioning deviation caused by measurement errors can be effectively reduced, and the situation that the traditional trilateral positioning algorithm fails is avoided.

Description

Technical Field

[0001] The present invention relates to the technical field of positioning algorithms, and particularly relates to a positioning method, a system, a vehicle, and an intelligent entry and start system. Background Art

[0002] Currently, in the automotive PEPS (Passive Entry & Passive Start, intelligent entry and start system), the solution based on Bluetooth Low Energy (BLE) has become the current mainstream technology due to its advantages such as good user experience and low cost. Among the current positioning algorithms applied in the Bluetooth Low Energy solution, the trilateration algorithm is the most widely used. However, due to interference factors such as environmental noise, reflection, and refraction, it will lead to large positioning errors or even algorithm failure. Summary of the Invention

[0003] In order to solve the technical problems in the related art, the present invention provides a positioning method, a system, a vehicle, and an intelligent entry and start system.

[0004] In order to achieve the above object, the technical solutions adopted by the present invention include: According to the first aspect of the present invention, a positioning method is provided, including the following steps: Step S1: Three anchor points obtain the positioning signals sent by an unknown node, calculate the corresponding RSSI values respectively, and calculate the distances between the three anchor points and the unknown node according to the corresponding RSSI values; Step S2: Draw circles with the three anchor points as the centers and the distances between the three anchor points and the unknown node as the radii respectively to obtain three circles, and judge the current state of the three circles according to the positional relationship of the three circles; Step S3: Perform weighted compensation on the current state of the three circles respectively to obtain the position of the unknown node.

[0005] Optionally, in the step S2, the current state of the three circles includes that the three circles have a common intersection point, the three circles intersect in a common area, the three circles intersect pairwise but have no common area, two circles do not intersect, one circle does not intersect with the other two circles, and the three circles are all separated; the step S3 specifically includes: For the case where the three circles have a common intersection point, use the least squares method to calculate the coordinates of the unknown node; For the case where the three circles intersect in a common area, use the centroid point of the common area as the unknown node; For the case where the three circles intersect pairwise but have no common area, use the centroid point of the area not occupied in the middle of the three circles as the unknown node; For two non-intersecting circles, the weighted compensation method is adopted to increase the radii of the two separated circles so that the two separated circles can be tangent after compensation. The compensation factor is the ratio of the distance between the centers of the two separated circles to the sum of their radii. Then, the centroid of the unoccupied area among the three circles after weighted compensation is taken as the unknown node. For one circle that does not intersect with the other two circles, the weighted compensation method is adopted to increase the radius of this circle so that it can intersect with the other two circles. The compensation factor is the maximum value among the ratios of the distances from this circle to the centers of the other two circles and the sum of their radii. Then, the centroid of the unoccupied area among the three circles after weighted compensation is taken as the unknown node. For three mutually separated circles, the weighted compensation method is adopted to increase the radii of the three circles so that the three circles intersect pairwise. The compensation factor is the maximum value among the ratios of the distances between the centers of every two circles and the sum of the radii of these two circles. Then, the centroid of the unoccupied area among the three circles after weighted compensation is taken as the unknown node.

[0006] Optionally, step S1 further includes: setting another anchor point and obtaining the positioning signal sent by the unknown node, calculating the corresponding RSSI value, and calculating the distance between this anchor point and the unknown node according to this RSSI value. Step S2 includes: respectively drawing circles with four anchor points as the centers and the distances between the four anchor points and the unknown node as the radii to obtain four circles, and combining any three of the four circles to obtain four different current three-circle states. Step S3 specifically includes: performing weighted compensation respectively according to the four different current three-circle states to obtain the preliminary coordinates of the four unknown nodes, then setting weights respectively according to the sum of the reciprocals of the radii of the circles with the anchor points as the centers in each combination, and using the set weights to perform weighted summation on the four preliminary coordinates to obtain the positioning coordinates of the unknown node.

[0007] Optionally, the positioning method further includes introducing the Kalman filtering algorithm to process the positioning signal, specifically including: Step S4-1-1: obtaining the position coordinates of the unknown node output by step S3, and judging whether it is the position coordinates of the unknown node obtained for the first time. If so, set it as the initial positioning coordinates of the unknown node. If not, enter step S4-1-2; Step S4-1-2: performing state prediction and covariance prediction in sequence; among them, the calculation formula for state prediction is: ; The calculation formula for covariance prediction is: ; In the formula, is the state estimate value of this prediction, is the state transition matrix, is the transpose matrix of the state transition matrix, is the state estimate obtained from the previous update, is the covariance matrix obtained from this update, is the covariance matrix obtained from the previous update, is the process noise covariance matrix; Step S4-1-3: Calculate the Kalman gain, state update, and covariance update in sequence; among them, the calculation formula for the Kalman gain is: The formula for state update is: The formula for covariance update is: In the formula, is the Kalman gain, is the observation matrix, is the observation matrix of the transposed matrix, is the measurement noise covariance matrix, is the state estimate at time k, is the measured value, is the state covariance matrix at time k, is the identity matrix; Step S4-1-4: Use the updated state value as the final position coordinate of the unknown node.

[0008] Optionally, the positioning method further includes introducing an extended Kalman filter algorithm to process the positioning signal, specifically including: Step S4-2-1: Obtain the position coordinate of the unknown node output in Step S3, and determine whether it is the position coordinate of the unknown node obtained for the first time. If so, set it as the initial positioning coordinate of the unknown node. If not, go to Step S4-2-2; Step S4-2-2: Perform state prediction and covariance prediction in sequence; among them, the calculation formula for state prediction is: ; The calculation formula for covariance prediction is: ; In the formula, is the state estimate of this prediction, is the non-linear state transition function, is the state estimate obtained from the previous update, is the covariance matrix obtained from this update, is the Jacobian matrix of the non-linear state transition function, is the covariance matrix obtained from the previous update, is the process noise covariance matrix; Step S4-2-3: Perform measurement residual, calculate residual covariance, calculate Kalman gain, state update, and covariance update in sequence; among them, the calculation formula for measurement residual is: The calculation formula for the residual covariance is: ; The calculation formula for the Kalman gain is: ; The formula for state update is: ; The formula for covariance update is: ; In the formula, is the difference between the measured value and the predicted value, is the measured value, is the predicted value, is the residual covariance, is the Jacobian matrix of the measurement function, is the transpose matrix of, is the uncertainty of the measurement noise, is the Kalman gain, is the state estimate value at time k, is the state covariance matrix at time k, is the identity matrix; Step S4-2-4: Use the updated state value as the final position coordinate of the unknown node.

[0009] According to the second aspect of the present invention, there is also provided a positioning system, which is applied to the positioning method described in any one of the technical solutions of the first aspect of the present invention. The positioning system includes: an unknown node, a data processing unit, and at least three anchor points. The unknown node is used to send a positioning signal; the anchor points are used to receive the positioning signal sent by the unknown node; the data processing unit is used to calculate the corresponding RSSI values respectively according to the positioning signal, and is used to calculate and calculate the distances between three anchor points and the unknown node respectively according to the corresponding RSSI values, and is used to judge the current three-circle state, and is used to perform weighted compensation on it respectively according to the current three-circle state to obtain the position of the unknown node.

[0010] According to the third aspect of the present invention, there is also provided a vehicle, which is characterized in that it includes the positioning system described in the second aspect of the present invention, and the anchor points and the data processing unit are arranged on the vehicle.

[0011] Optionally, the vehicle further includes a smart key, and the unknown node is arranged on the smart key.

[0012] According to a fourth aspect of the present invention, an intelligent access and start system is further provided, which includes an unknown node, a data processing unit, and at least three anchor points. The unknown node is configured to be set on a mobile unlocking terminal matched with a vehicle and to send a positioning signal; the anchor points are configured to be set on the vehicle and to receive the positioning signal sent by the unknown node; the data processing unit is configured to be set on the vehicle and to execute the positioning method described in any one of the technical solutions of the first aspect of the present invention.

[0013] Beneficial effects: 1. The positioning method provided by the present invention is improved based on the traditional trilateration algorithm, and corresponding compensation is performed according to different three-circle position states, so that the positioning deviation caused by measurement errors can be effectively reduced, and the situation where the traditional trilateration algorithm fails can be avoided.

[0014] 2. Other beneficial effects or advantages of the present invention will be described in detail in the specific implementation manners. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments.

[0016] Among them: Figure 1 is a schematic flowchart of the steps of a positioning method provided by an exemplary embodiment; Figure 2 is a schematic diagram of three circles intersecting pairwise and having a common intersection point; Figure 3 is a schematic diagram of three circles intersecting in a common area; Figure 4 is a schematic diagram of three circles intersecting pairwise but having no common area; Figure 5 is a schematic diagram of two non-intersecting circles among three circles; Figure 6 is a schematic diagram of one circle not intersecting with the other two circles; Figure 7 is a schematic diagram of three circles being mutually separated; Figure 8 is a schematic flowchart of the steps of a weighted compensation algorithm provided by an exemplary embodiment; Figure 9 is a schematic diagram of the positioning principle of four anchor points; Figure 10 is a schematic flowchart of the steps of introducing KF provided by an exemplary embodiment; Figure 11 is a schematic flowchart of the steps of introducing EKF provided by an exemplary embodiment; Figure 12It is a schematic diagram of the layout of signal transceiver devices during the actual measurement of the static positioning of the three-anchor-point algorithm; Figure 13 It is a schematic diagram of the layout of signal transceiver devices during the actual measurement of the static positioning of the four-anchor-point algorithm; Figure 14 It is a schematic diagram of the layout of reference trajectory 1; Figure 15 It is a schematic diagram of the layout of reference trajectory 2; Figure 16 It is a schematic diagram of the layout of reference trajectory 3. Detailed implementation manner

[0017] For the relevant technical personnel to have a clearer and more accurate understanding of the technical solution of the present invention, the problems existing in the existing relevant technologies will be described in more detail below.

[0018] Please refer to Figure 2 , the principle of the trilateration method is: three anchor points (A, B, and C) are arranged in advance. Subsequently, starting from the unknown node (D) (i.e., the unknown node sends a positioning signal), the distances to these three anchor points are measured respectively, and multiple circles are drawn with these distances as the radii. In an ideal situation, these circles will intersect at the same point (i.e., the estimated position of the unknown node). Assuming the coordinates of anchor points A, B, and C are ( ), ( ), ( ), and the coordinates of the unknown node D are ( ), then the distances between the three anchor points and the unknown node are respectively , and .

[0019] At this time, with , and as the radii and anchor points A, B, and C as the centers, the three circles intersect at point D, that is: The solution of this equation is the estimated coordinates of the unknown node D. Expanding the equation, we can get: Subtracting the first two equations in this formula from the third equation respectively, we can get: The coordinates of the unknown node D can be obtained as: However, during the ranging stage, interference factors such as noise, reflection and refraction in the environment may cause large errors in the RSSI (Signal Received Strength Indicator) value, which in turn leads to a deviation between the calculated distance value and the actual distance value. When the calculated distance value with large error is used to infer the coordinates, the three-sided positioning algorithm may have a positioning error that is too large, and may even cause the algorithm to fail.

[0020] Therefore, the present invention provides a brand-new solution, that is, the positioning method of the present invention. The technical idea of ​​the positioning method of the present invention is: to correct and optimize the different position states of the three circles in the calculation process of the three-side positioning algorithm respectively, so as to reduce the positioning error and avoid the problem of algorithm failure.

[0021] The technical solution of the present invention is described in detail below with reference to the accompanying drawings.

[0022] In order to solve the problem that the existing three-sided positioning algorithm has a large positioning error and may fail to work, the present invention proposes a positioning method. The positioning method of the present invention considers various three-circle intersection situations (i.e., three-circle position states) such as Figures 2 to 7 shown.

[0023] In the present invention, the weighted compensation processing method for each three-circle intersection situation includes: 1) For the case where the three circles have a common intersection, that is, Figure 2 shown.

[0024] This situation is an ideal situation (i.e., the three circles intersect at a common point), and no correction is required. According to the three-sided positioning algorithm, the matrix can be set , and as follows: The formula of the three-side positioning algorithm can be simplified as follows: ; Using the least squares method, the coordinates of point D can be solved as: .

[0025] 2) For the case where the three circles intersect in a public area, that is, Figure 3 shown.

[0026] When the measurement error is small, and the three circles intersect each other but do not intersect at one point but produce a common area, the same method as the triangle centroid method can be used to take the centroid point D of the common area as the position of the target to be measured.

[0027] Among them, Figure 3As shown in the figure, circles A and B intersect at points F1 and F2. When selecting the intersection points, it can be judged based on the distances from points F1 and F2 to the center of circle C. Select the closer point F1 and discard the farther point F2; similarly, for the intersection points G1 and G2 of circles B and C, select point G1; for the intersection points E1 and E2 of circles A and C, select point E1. In this way, the coordinates of the centroid point of the common area composed of points F1, E1, and G1 can be used to obtain the coordinates of point D. The coordinates of the three intersection points are F1( ), G1( ), E1( ), then the coordinates of point D are: ; 3) For the case where the three circles intersect pairwise but have no common area, as shown in Figure 4 the figure.

[0028] When the ranging error causes the three circles to intersect pairwise but have no common area, its positioning principle is the same as the case where the three circles intersect at a common area, that is, take the centroid point of the area not occupied in the middle of the three circles composed of points F1, G1, and E1 as the position of the target to be measured.

[0029] 4) For the case where two of the three circles do not intersect, as shown in Figure 5 the figure.

[0030] When the ranging error causes one of the three circles to be separated from any one of the other two circles, the traditional trilateration algorithm will fail and the coordinate calculation of the unknown node cannot be performed. In the present invention, a weighted compensation method can be considered to increase the radii of the separated two circles for error compensation. The compensation factor k is the ratio of the distance between the centers of the non-intersecting two circles to the sum of the radii.

[0031] Circles B and C do not intersect, and the distance between the centers of circles B and C is . Denote the radii of circles B and C after compensation as and , then the formula for weighted compensation can be: In this way, circles B and C after weighted compensation can be tangent to each other at point G. At this time, it is the same as the case where the three circles intersect pairwise but have no common area, that is, the centroid point of the area composed of points F1, G, and E1 can be taken as the position of the target to be measured.

[0032] 5) For the case where one circle does not intersect with the other two circles, as shown in Figure 6 the figure.

[0033] When the ranging error causes any one of the three circles not to intersect with the other two circles, the traditional trilateration algorithm will fail and the coordinate calculation of the unknown node cannot be performed. In the present invention, the circle can be made to intersect with the other two circles by means of weighted compensation, and the compensation factor k takes the larger value among the ratios of the distances between the centers of the circle and the sum of the radii of the other two circles.

[0034] As Figure 6 shown, circle B does not intersect with circles A and C, and the distances between the centers of circle B and circles A and C are respectively and . If the ratio of the distance between the centers and the sum of the radii satisfies the following relationship: ; that is, the ratio of the distance between the centers of circles A and B and the sum of their radii is larger, then weighted compensation can be performed on circles A and B. Denote the radii of circles A and B after compensation as and respectively, and the formula for weighted compensation can be expressed as: In this way, after weighted compensation, circles A and B will be tangent to each other at point F. At the same time, circles B and C intersect at points G1 and G2. At this time, it is the same as the case where the three circles intersect pairwise but there is no common area, that is, the centroid point D of the area composed of points F, G1, and E1 can be taken as the position of the target to be measured.

[0035] 6), For the case where all three circles are separated from each other, as Figure 7 shown.

[0036] When the ranging error is too large, it will cause the situation where all three circles are separated from each other, and the traditional trilateration algorithm will fail and the coordinate calculation of the unknown node cannot be performed. In the present invention, the radii of the three circles can be increased by weighted compensation so that the three circles intersect pairwise, and the compensation factor k takes the largest value among the ratios of the distances between the centers of every two circles and the sum of the radii of these two circles.

[0037] Circles A, B, and C do not intersect with each other, and the distances between the centers of circles A, B, and C pairwise are respectively , and . If the ratio of the distance between the centers and the sum of the radii satisfies the following relationship: ; that is, the ratio of the distance between the centers of circles A and B and the sum of their radii is the largest. Then weighted compensation can be performed on all three circles A, B, and C. Denote the radii of the three circles after compensation as , and respectively, and the formula for weighted compensation can be expressed as: In this way, after weighted compensation, circles A and B will be tangent to each other at point F. At this time, it is the same as the case where the three circles intersect pairwise but have no common area. That is, the centroid point D of the area formed by points F, G1, and E1 can be taken as the position of the target to be measured.

[0038] The process of the weighted compensation algorithm of the present invention can be referred to Figure 8 as shown.

[0039] Based on the above content, in an exemplary embodiment, the positioning method of the present invention may include the following steps: Step S1: Three anchor points obtain the positioning signals sent by the unknown node, calculate the corresponding RSSI values respectively, and calculate the distances between the three anchor points and the unknown node according to the corresponding RSSI values; Step S2: Draw circles with the three anchor points as the centers and the distances between the three anchor points and the unknown node as the radii respectively to obtain three circles, and judge the current state of the three circles according to the positional relationship of the three circles; Step S3: Perform weighted compensation on the current state of the three circles respectively to obtain the position of the unknown node.

[0040] Through the above technical solution, the positioning method provided by the present invention is improved on the basis of the traditional trilateration algorithm, and corresponding compensation is performed according to different positional states of the three circles, so that the positioning deviation caused by measurement errors can be effectively reduced, and the situation where the traditional trilateration algorithm fails can be avoided.

[0041] As can be seen from the above content, the current weighted compensation trilateration algorithm relies on three anchor points to collect RSSI data. However, this algorithm also has disadvantages. For example, when the RSSI value measured by a certain anchor point has a large deviation, the overall positioning accuracy is insufficient to meet the requirements.

[0042] Based on this, in an exemplary embodiment of the present invention, on the basis of the three anchor points of the original trilateration algorithm, a new anchor point is additionally added, and the combination scheme of four anchor points combined with weighted compensation is used to further improve the positioning accuracy and fault tolerance.

[0043] In the present invention, the principle of the combination scheme of four anchor points combined with weighted compensation is as follows: First, any three of the four anchor points are combined to obtain four different anchor point combinations. Then, the weighted compensation trilateration algorithm is applied to these four different anchor point combinations respectively to calculate four preliminary coordinates. Subsequently, weights are set respectively according to the sum of the reciprocals of the radii of the circles with the anchor points as the centers in each anchor point combination, and these weights are used to perform weighted summation on the four preliminary coordinates to obtain the final positioning coordinates. The positioning principle can be referred to Figure 9 as shown.

[0044] As shown Figure 9 , the radii of the four anchor circles A, B, C, and D are respectively , , and . The four circles are combined in groups of any three, divided into four groups: (A, B, C), (A, B, D), (A, C, D), and (B, C, D). A preliminary coordinate is obtained for each group of three anchor circles through the weighted compensation trilateral positioning algorithm. The four preliminary coordinates obtained through calculation are respectively , , and . The weight of each group of preliminary coordinates is determined by the radii of its three anchor circles, that is, the smaller the radius, the greater the weight, meaning that the influence of the anchor point on the final result is greater. The specific weights can be respectively: The calculated preliminary coordinates of each group are weighted and accumulated according to their corresponding weights, and the weighted coordinates obtained are: At the same time, the weights are accumulated to obtain the total weight as: The weighted coordinates are normalized according to the total weight, and the final positioning point coordinates obtained are: .

[0045] In this way, compared with the positioning method provided in the foregoing embodiment, this embodiment can effectively reduce the influence of single-group errors on the result through the combined calculation of multiple groups of anchor points.

[0046] In addition, in this embodiment, the reciprocal of the radius is used as the weight, that is, the nearby anchor points are given priority, reducing the interference of the distant anchor points, and it is applicable to anchor point groups with arbitrary distributions.

[0047] Based on the above content, in an exemplary embodiment, the positioning method of the present invention may further include the following steps: Step S1: Four anchor points acquire the positioning signal sent by the unknown node, calculate the corresponding RSSI value, and calculate the distance between the anchor point and the unknown node according to the RSSI value; Step S2: Respectively draw circles with the four anchor points as the centers and the distances between the four anchor points and the unknown node as the radii to obtain four circles, and combine any three of the four circles to obtain four different current three-circle states; Step S3: Perform weighted compensation according to four different current three-circle states to obtain the preliminary coordinates of the four unknown nodes. Then, set weights according to the sum of the reciprocals of the radii of the circles centered on the anchor points in each combination, and use the set weights to perform weighted summation on the four preliminary coordinates to obtain the positioning coordinates of the unknown nodes.

[0048] Although the combined scheme of the above four anchor points combined with weighted compensation can effectively reduce the influence of RSSI errors, however, in actual application scenarios, this combined scheme may still be affected by various noises and interferences, resulting in fluctuations or deviations in the positioning results.

[0049] For example, in the positioning application of PEPS, the system needs to capture and respond to the dynamic changes of the position of the target to be measured in real time and accurately. However, the combined scheme of four anchor points combined with weighted compensation mainly relies on the RSSI measurement values in a static environment for calculation, which may lead to difficulties in real-time processing, making it difficult for this algorithm to quickly adapt to dynamic changes, and thus resulting in inaccurate positioning results.

[0050] Therefore, how to improve the real-time performance and accuracy of the positioning results becomes another problem to be solved.

[0051] In an embodiment of the present invention, on the basis of the combined scheme of four anchor points combined with weighted compensation, Kalman Filter (KF) and Extended Kalman Filter (EKF) are respectively introduced for algorithm optimization to improve the real-time performance and accuracy of the positioning results.

[0052] Specifically, please refer to Figure 10 As shown, for the algorithm introducing Kalman Filter (KF), it consists of two main stages, namely the prediction stage and the update stage. In the prediction stage, based on the state model of the system, a prior estimate of the state and its uncertainty at the current moment is predicted, which is divided into two steps: state prediction and covariance prediction. The formula for state prediction is as follows: ; where is the state transition matrix, which describes the change of the system state from one moment to the next moment, is the state estimate obtained from the previous update, is the state estimate value of this prediction. Its function is to predict the state at the current moment according to the state transition model of the system . The formula for covariance prediction is as follows: ; where is the covariance matrix obtained from the previous update, representing the uncertainty of the previous state estimate. is the additional uncertainty introduced after the state transition, is the transpose matrix of the state transition matrix, is the process noise covariance matrix, which describes the impact of random noise in the system on the state. Its role is to estimate the uncertainty of the state at the current moment based on the predicted state equation.

[0053] In the update stage, the prediction result is updated by combining measurement data to obtain a more accurate posterior state estimate, which is divided into three steps: calculating the Kalman gain, state update, and covariance update. The calculation formula of the Kalman gain is as follows: ; where, is the Kalman gain, is the predicted uncertainty, is the observation matrix, which maps the state to the measurement space, is the observation matrix of the transpose matrix, is the measurement noise covariance matrix, which describes the uncertainty of the measurement data. Its role is to calculate the Kalman gain for adjusting the weight between the predicted value and the measured value. The formula for state update is as follows: ; where, is the state estimate value at time k, is the measured value, is the residual between the measured value and the predicted value, is to correct the predicted value with the residual to approach the measured value. The formula for covariance update is as follows: ; where, is the state covariance matrix at time k, is to correct the impact of the measurement on the state estimate, is the identity matrix, is the correction matrix, which is used to reduce the uncertainty of the state.

[0054] Please refer to Figure 11 As shown, for the algorithm introducing the Extended Kalman Filter (EKF), it also has two main stages: the prediction stage and the update stage. EKF calculates the prior estimates of the state and covariance in the prediction stage. The formula for state prediction is as follows: ; its role is to use the state estimate at the previous moment and the nonlinear state transition function to predict the state

[0055] at the current moment, which represents the state estimate of the system without combining the measured value. ; where, is the Jacobian matrix of the state transition function, which is used to linearize the nonlinear function ; is the covariance matrix obtained from the previous update, The covariance matrix obtained for this update quantifies the uncertainty predicted at the current moment and includes the uncertainty of the system itself (through the process noise ), as well as the uncertainty of the state at the previous moment.

[0056] In the update stage, the EKF combines the measurement values to correct the predicted values. Different from the KF, the EKF is divided into five steps in the update stage: measurement residual, calculation of residual covariance, calculation of Kalman gain, state update, and covariance update. The formula for the measurement residual is: ; its role is to calculate the difference between the measurement value and the predicted value , which is called the measurement residual or innovation and is used to correct the predicted state. The formula for the residual covariance is: ; where is the Jacobian matrix of the measurement function, which is used to linearize the non-linear measurement function , is 's transpose matrix. Its role is to calculate the residual covariance , which includes the uncertainty of the measurement noise and the contribution of the predicted covariance , reflecting the uncertainty magnitude of the measurement value and the predicted value. The formula for the Kalman gain is: ; where is higher, the greater the influence of the measurement value; is lower, the greater the influence of the predicted state. The formula for the state update is: ; its role is to correct the predicted state according to the measurement value to obtain the updated state . The formula for the covariance update is: ; where is the state covariance matrix at time k, is the identity matrix. Its role is to update the state covariance matrix, reduce the uncertainty of the predicted state, and reflect the accuracy of the updated state.

[0057] It can be understood that the prediction and update processes are written as predict and update methods respectively. First, the coordinates of the next moment are predicted through the predict method, and then the predicted coordinates are updated using the newly obtained coordinates through the update method. During the positioning process, whenever new coordinates estimated by the four-anchor weighted compensation trilateration positioning algorithm are available, the KF or EKF updates the positioning through two steps.

[0058] Among them, refer to Figure 10 , when the KF is introduced, the steps are as follows: 1). First positioning: When the positioning information is obtained for the first time, no filtering is performed, and only a default initial positioning coordinate is given.

[0059] 2), Update and prediction: If positioning information has been obtained before, every time new coordinates are acquired, the KF filtering process will be entered. First, based on the state transition matrix A and the previous state x, the predicted coordinates at the current moment are generated. At this time, the predicted state will be updated using the state vector x and covariance matrix P at the previous moment to calculate the estimated state at the current moment. Then, the new positioning measurement coordinates are defined as z, and the filtering gain K is calculated, which is a coefficient that weighs the predicted coordinates and the newly acquired coordinates. Finally, combining the predicted coordinates and the newly acquired coordinates, the state vector x is updated, and the covariance matrix P is updated to reduce uncertainty.

[0060] Refer to Figure 11 , when introducing EKF, the steps are as follows: 1), First positioning: When obtaining positioning information for the first time, no filtering is performed, and only a default initial positioning coordinate is given. At the same time, the state vector of EKF is initialized, and the initial position information is assigned to the state vector x[:2].

[0061] (2) Update and prediction: If positioning information has been obtained before, every time new coordinates are acquired, the EKF filtering process will be entered. First, based on the state transition function f(x) (non-linear) and the state transition Jacobian matrix jacobian_F(x), the predicted coordinates are calculated. In this process, using the state vector x and covariance matrix P at the previous moment, combined with the process noise covariance Q, the predicted state vector x and the state covariance matrix P are updated. Then, the new positioning measurement coordinates are defined as z, and using the measurement equation h(x) (non-linear) and the observation matrix Jacobian jacobian_H(x), the measurement residual y = z - h(x) is calculated. Through the measurement residual y and the filtering gain K, the state vector x and the state covariance matrix P are updated to reduce the uncertainty of prediction and observation.

[0062] To verify the positioning effect of the improved algorithm in the real PEPS positioning scenario, the present invention conducts on-site actual measurements and data analysis on the algorithms before and after improvement.

[0063] The positioning measurements were carried out indoors and outdoors respectively, and the positioning was carried out by Bluetooth RSSI ranging. When the Bluetooth signal transceiver is used as the signal transmitter, its transmission power is set to 0 dBm, and the signal transmission interval is 30 ms. When it is used as the signal receiver, after collecting 30 RSSI data, a Gaussian combined mean filtering process is performed to obtain an RSSI value. The filtering process is directly implemented on the signal receiver hardware to reduce the computational burden on the PC side and speed up the filtering speed. The filtered RSSI values are transmitted to the PC side through serial communication, and the positioning calculation is performed by the GUI upper computer software developed based on PyQt5, and the positioning effect of the algorithm is visually displayed. Through the GUI upper computer software, the parameter values of the path loss model can be configured and the position coordinates of the signal receiving anchor points and the signal transmitting positioning reference points can be set, so as to realize the positioning function.

[0064] 1. Static positioning measurement To verify the performance of the improved algorithm in practical applications, positioning measurements and comparative analysis were carried out by comparing it with the traditional trilateration method. In the experiment, to simulate the positioning scenario of PEPS, three signal receivers were used as vehicle body anchor points, and their coordinates were set to (1000, 0), (0, -1000), and (-1000, 0) respectively, with the unit being unified as millimeters (mm). Then, a signal transmitter was used to simulate the car key and was placed at four positioning reference positions in turn. During the whole experiment, the transceiver devices were kept at the same height, and the GUI upper computer was used to verify and compare the positioning effect of the algorithm. The layout of the signal transceiver devices is as Figure 12 shown.

[0065] As Figure 12 shown, the coordinates of the four positioning reference points are reference point 1 (1500, 0), reference point 2 (0, -1500), reference point 3 (-500, -500), and reference point 4 (500, 500). These four reference points cover multiple positions inside the triangle, outside but close to the edge, and on one side and in the vertical direction of the anchor points, and can comprehensively evaluate the positioning accuracy of the method in different regions. When conducting the positioning measurement, 30 positioning data were predicted and recorded at each positioning reference point to evaluate the performance of the algorithm.

[0066] (1) Measurement of the traditional trilateration algorithm

[0067] To evaluate the positioning performance of the traditional trilateration algorithm and compare it with the improved algorithm, positioning measurements were carried out on the four reference points respectively. By analyzing the coordinate data of the reference points and the predicted points for indoor and outdoor positioning, various parameters for evaluating the positioning performance of this algorithm can be calculated, as shown in Table 1 below.

[0068] Table 1 Performance evaluation parameters of the traditional trilateration algorithm (unit: mm) As can be seen from Table 1 above, whether in indoor or outdoor environments, the mean errors at different reference points are relatively large, indicating the instability of the traditional trilateration positioning algorithm during the positioning process. In addition, the generally high error variances reveal the wide distribution of the positioning point deviations, thus proving the discreteness of the predicted positioning points, and this is further confirmed by the maximum and minimum errors. To sum up, there are significant defects in the traditional trilateration positioning algorithm, with large discreteness of the predicted positioning points and very unsatisfactory positioning effects.

[0069] (2)Actual measurement of the weighted compensation trilateration positioning algorithm provided by an exemplary embodiment of the present invention.

[0070] To verify the positioning effect of the improved weighted compensation trilateration positioning algorithm of the present invention, tests were conducted under the same actual measurement conditions as the traditional trilateration positioning algorithm. The indoor and outdoor positioning actual measurement performance indicators of the weighted compensation trilateration positioning algorithm are shown in Table 2 below.

[0071] Table 2 Performance evaluation parameters of the weighted compensation trilateration positioning algorithm (unit: mm) As can be seen from Table 2 above, the numerical values of the performance parameters of the weighted compensation trilateration positioning algorithm of the present invention are generally lower than those of the traditional trilateration positioning algorithm, which proves its better positioning effect. Especially in the key indicator of error variance, the improved algorithm has achieved a significant decrease in the parameter value, which directly reflects that its predicted positioning points are more concentrated and the degree of discreteness has been effectively controlled.

[0072] However, it can be understood that although this algorithm has been improved on the basis of the traditional trilateration positioning algorithm, there are still some deficiencies in its actual measurement effect.

[0073] (3)Actual measurement of the four-anchor-point weighted compensation trilateration positioning algorithm.

[0074] To further improve the positioning accuracy, optimization is considered on the basis of the weighted compensation trilateration positioning algorithm. The current weighted compensation trilateration positioning algorithm relies on three signal receivers as vehicle body anchor points to collect RSSI data. However, this algorithm also has drawbacks, that is, the RSSI value measured by a certain anchor point may have a large deviation, resulting in inaccurate positioning. To reduce the fault tolerance problem caused by this deviation, a new vehicle body anchor point with coordinates (0, 1000) is added on the basis of the original three anchor points, and the weighted compensation trilateration positioning is realized through four anchor points. To compare the algorithm effects, the number and positions of the positioning reference points remain unchanged, and its signal transceiver device layout is as Figure 13 shown. Under the same actual measurement conditions as before, positioning simulation was carried out on the optimized four-anchor-point weighted compensation trilateration positioning algorithm. The indoor and outdoor positioning actual measurement performance indicators of the four-anchor-point weighted compensation trilateration positioning algorithm are shown in Table 3 below.

[0075] Table 3 Performance Evaluation Parameters of the Four-Anchor-Point Weighted Compensation Trilateration Algorithm (Unit: mm) As can be seen from Table 3 above, in terms of the mean error index, the four-anchor-point weighted compensation trilateration algorithm of the present invention performs better than the weighted compensation trilateration algorithm in both indoor and outdoor environments and at the four positioning points. Secondly, the error variance of this algorithm and the gap between the maximum error and the minimum error are also generally lower. These conclusions indicate that the four-anchor-point weighted compensation trilateration algorithm has higher positioning accuracy, and the distribution of its positioning points is more concentrated and stable.

[0076] (4) Actual measurement of the four-anchor-point weighted compensation trilateration + KF / EKF algorithm.

[0077] Under the same actual measurement conditions, KF is introduced into the four-anchor-point weighted compensation trilateration algorithm for positioning simulation comparison. The indoor and outdoor positioning actual measurement performance indicators of the four-anchor-point weighted compensation trilateration + KF / EKF algorithm are shown in Table 4 and Table 5 below respectively.

[0078] Table 4 Performance Evaluation Parameters of the Four-Anchor-Point Weighted Compensation Trilateration + KF Algorithm (Unit: mm) Table 5 Performance Evaluation Parameters of the Four-Anchor-Point Weighted Compensation Trilateration + EKF Algorithm (Unit: mm) As can be seen from the comparison in Table 4 and Table 5 above, whether in the indoor or outdoor environment, after adding KF or EKF to the four-anchor-point weighted compensation trilateration algorithm, the mean error, error variance, and maximum error of each reference point are significantly reduced, indicating that introducing the KF or EKF algorithm can effectively improve the positioning accuracy.

[0079] According to the data in Table 4 and Table 5 above, the positioning parameters of the four-anchor-point weighted compensation trilateration + KF / EKF algorithm in the outdoor environment are generally better than those in the indoor environment. Because the outdoor environment is relatively open, the signal propagation path is relatively direct, and it is less affected by factors such as multipath effects, thus improving the positioning accuracy. Comparing these two algorithms, the four-anchor-point weighted compensation trilateration + KF algorithm is better. Especially, the value of the error variance parameter of it is smaller than that of the four-anchor-point weighted compensation trilateration + EKF algorithm, making its positioning prediction points more accurate and concentrated. Although both KF and EKF can correct and predict the positioning, EKF is more effective in dealing with nonlinear problems. However, in the four-anchor-point weighted compensation trilateration algorithm, the performance of introducing KF is better, because the nonlinear degree of this positioning problem is not high, and the simplicity and robustness of KF make it more suitable for this application scenario.

[0080] The average values of the performance evaluation parameters of the four positioning reference points obtained by measuring each algorithm indoors and outdoors are shown in Table 6 below.

[0081] Table 6 Mean values of performance evaluation parameters of each algorithm (unit: mm) Compared with the traditional trilateration positioning algorithm, in the indoor environment, the mean error and error variance of the weighted compensation trilateration positioning algorithm are reduced by 31.44% and 55.54% respectively, the four-anchor weighted compensation trilateration positioning algorithm is reduced by 48.92% and 79.45%, the four-anchor weighted compensation trilateration positioning + EKF algorithm is reduced by 76.90% and 93.01%, and the four-anchor weighted compensation trilateration positioning + KF algorithm is reduced by 84.01% and 97.48%. In the outdoor environment, the positioning performance of each algorithm has been improved. Compared with the traditional trilateration positioning algorithm, the mean error and error variance of the weighted compensation trilateration positioning algorithm are reduced by 50.00% and 60.75% respectively, the four-anchor weighted compensation trilateration positioning algorithm is reduced by 62.06% and 81.24%, the four-anchor weighted compensation trilateration positioning + EKF algorithm is reduced by 89.68% and 98.57%, and the four-anchor weighted compensation trilateration positioning + KF algorithm is reduced by 92.37% and 99.35%.

[0082] In summary, the four-anchor weighted compensation trilateration positioning + KF algorithm shows better positioning performance both indoors and outdoors.

[0083] 2. Dynamic positioning measurement To more accurately simulate the PEPS positioning scenario, in addition to static positioning, dynamic positioning measurement is also required to test the positioning performance of the algorithm. Since PEPS requires its positioning system to accurately identify the position changes of people approaching the vehicle from a distance and near the vehicle, three dynamic positioning reference trajectories are designed, and the positioning performance of the positioning algorithm on the reference trajectories is evaluated through actual measurement. The three dynamic positioning reference trajectories are as Figures 14 to 16 shown, where Figure 14 is the layout schematic diagram of reference trajectory 1, Figure 15 is the layout schematic diagram of reference trajectory 2, Figure 16 is the layout schematic diagram of reference trajectory 3.

[0084] At Figures 14 to 16Among them, the thick line represents the reference trajectory, and the arrow on it indicates the traveling direction. Multiple positioning reference points are set on each reference trajectory, and these points serve as the key reference benchmarks for the advancement of the dynamic trajectory. During the positioning measurement, each time the tester starts from the positioning reference point 1 with a signal transmitter simulating a car key and travels to each subsequent positioning reference point one by one at a constant speed until reaching the last positioning reference point. At the start and during the measurement process, whenever at any positioning reference point, an appropriate stop is made and the predicted position of the current positioning reference point is recorded for positioning error analysis. For reference trajectory 1, it starts from positioning reference point 1 and ends until it returns to this reference point again. For reference trajectories 2 and 3, they both start from positioning reference point 1 and end at positioning reference point 3.

[0085] During the measurement stage, based on the results of the previous static positioning tests, the three algorithms with the best performance, namely weighted compensation trilateration with four anchor points, weighted compensation trilateration with four anchor points + EKF, and weighted compensation trilateration with four anchor points + KF, were selected for indoor and outdoor dynamic positioning tests.

[0086] (1) Positioning measurement of reference trajectory 1.

[0087] Taking the positioning effects of each reference point in reference trajectory 1 as the dynamic positioning performance indicators of the algorithm, as shown in Table 7 below.

[0088] Table 7 Dynamic positioning performance evaluation parameters of three algorithms for reference trajectory 1 (unit: mm) Compared with the weighted compensation trilateration algorithm with four anchor points, in the indoor environment, the mean error and error variance of the weighted compensation trilateration with four anchor points + EKF algorithm are reduced by 51.66% and 77.38% respectively, and the weighted compensation trilateration with four anchor points + KF algorithm is reduced by 67.64% and 88.57%. And the positioning performance of each algorithm in the outdoor environment has been improved. Compared with the weighted compensation trilateration algorithm with four anchor points, the mean error and error variance of the weighted compensation trilateration with four anchor points + EKF algorithm are reduced by 70.19% and 91.77%, and the weighted compensation trilateration with four anchor points + KF algorithm is reduced by 78.65% and 96.05%.

[0089] (2) Positioning measurement of reference trajectory 2.

[0090] Taking the positioning effects of each reference point in reference trajectory 2 as the dynamic positioning performance indicators of the algorithm, as shown in Table 8 below.

[0091] Table 8 Dynamic positioning performance evaluation parameters of three algorithms for reference trajectory 2 (unit: mm) Compared with the four-anchor weighted compensation trilateration algorithm, in the indoor environment, the mean error and error variance of the four-anchor weighted compensation trilateration + EKF algorithm are reduced by 52.64% and 77.33% respectively, and the four-anchor weighted compensation trilateration + KF algorithm is reduced by 66.02% and 88.88%. And the positioning performance of each algorithm is improved in the outdoor environment. Compared with the four-anchor weighted compensation trilateration algorithm, the mean error and error variance of the four-anchor weighted compensation trilateration + EKF algorithm are reduced by 70.54% and 88.18%, and the four-anchor weighted compensation trilateration + KF algorithm is reduced by 79.84% and 95.70%.

[0092] (3)Actual measurement of reference trajectory 3 positioning.

[0093] Taking the positioning effect of each reference point in reference trajectory 3 as the dynamic positioning performance index of the algorithm, as shown in Table 9 below.

[0094] Table 9 Dynamic positioning performance evaluation parameters of three algorithms for reference trajectory 3 (unit: mm) Compared with the four-anchor weighted compensation trilateration algorithm, in the indoor environment, the mean error and error variance of the four-anchor weighted compensation trilateration + EKF algorithm are reduced by 50.81% and 76.90% respectively, and the four-anchor weighted compensation trilateration + KF algorithm is reduced by 64.14% and 90.08%. And the positioning performance of each algorithm is improved in the outdoor environment. Compared with the four-anchor weighted compensation trilateration algorithm, the mean error and error variance of the four-anchor weighted compensation trilateration + EKF algorithm are reduced by 72.26% and 91.44%, and the four-anchor weighted compensation trilateration + KF algorithm is reduced by 81.82% and 95.59%.

[0095] It can be seen from the above embodiments that: First, the four-anchor weighted compensation trilateration algorithm proposed by the present invention is improved on the basis of the traditional trilateration algorithm. By using the method of compensating the radius of the circle, the situation where the circles cannot intersect in trilateration is solved, thereby effectively reducing the positioning deviation caused by ranging errors. At the same time, on the basis of the traditional use of three anchors for positioning, a new anchor is added, which improves the fault tolerance rate of the positioning system when errors exist and further improves the positioning accuracy. Compared with the traditional trilateration algorithm, the mean errors of the four-anchor weighted compensation trilateration algorithm in indoor and outdoor static positioning actual measurements are reduced by 48.92% and 62.06% respectively.

[0096] Second, in the four-anchor weighted compensation trilateration positioning algorithm of the present invention, the Kalman filter and the extended Kalman filter algorithm are introduced for actual measurement analysis and comparison of positioning. Compared with the traditional trilateration positioning algorithm, the mean static positioning errors of the four-anchor weighted compensation trilateration + EKF algorithm indoors and outdoors are reduced by 76.90% and 89.68% respectively, and those of the four-anchor weighted compensation trilateration + KF algorithm are reduced by 84.01% and 92.37% respectively. Therefore, the Kalman filter is selected to optimize the algorithm. By introducing the Kalman filter, not only the ranging error is effectively reduced, making the predicted position of the positioning more accurate and concentrated, but also the performance and performance of the algorithm in the real-time dynamic positioning scenario are improved. After actual measurement verification of dynamic positioning on different reference trajectories, the results show that the four-anchor weighted compensation trilateration + KF algorithm can accurately track the position of the target to be measured and meet the positioning requirements of the automotive PEPS system.

[0097] According to the second aspect of the present invention, a positioning system is further provided, which is applied to the positioning method of any one of the technical solutions in the first aspect of the present invention. The positioning system includes an unknown node, a data processing unit, and at least three anchor points. Among them, the unknown node is used to send a positioning signal; the anchor point is used to receive the positioning signal sent by the unknown node; the data processing unit is used to calculate the corresponding RSSI value according to the positioning signal, and is used to calculate and calculate the distances between the three anchor points and the unknown node according to the corresponding RSSI value, and is used to judge the current three-circle state, and is used to perform weighted compensation on it according to the current three-circle state to obtain the position of the unknown node.

[0098] Through the above technical solution, the positioning system of the present invention performs corresponding compensation according to different three-circle position states, so that the positioning deviation caused by measurement errors can be effectively reduced, and the situation of the failure of the traditional trilateration positioning algorithm can be avoided. Moreover, the positioning method of the present invention has high accuracy and real-time performance, and can meet the positioning requirements of the automotive PEPS system.

[0099] According to the third aspect of the present invention, a vehicle is further provided, which is characterized in that it includes the positioning system of the second aspect of the present invention, and the anchor point and the data processing unit are arranged on the vehicle.

[0100] In this way, the vehicle of the present invention can achieve fast and accurate positioning in a variety of application scenarios when equipped with the positioning system of the present invention.

[0101] In an embodiment of the present invention, the vehicle may further include a smart key, and the unknown node is arranged on the smart key.

[0102] In this way, when the vehicle of the present invention is equipped with the positioning system of the present invention and the unknown node capable of sending positioning signals is equipped on the intelligent key, it can achieve fast and accurate positioning in a variety of application scenarios, and then facilitate the realization of other functions, such as keyless entry, or keyless start, etc.

[0103] It can be understood that, first, the vehicle of the present invention should be understood in a broad sense, which refers to all land mobile devices used for transporting people, goods or for other special purposes, generally driven by a power device and having wheels or tracks to be able to travel on roads or specific sites. Specifically, the vehicle of the present invention includes but is not limited to: passenger vehicles (sedans, sports cars, SUVs, MPVs, etc.), commercial vehicles (trucks, buses, special vehicles, such as light trucks, buses, fire trucks, garbage trucks, etc.), motorcycles, special vehicles (agricultural machinery), rail transit vehicles, modified vehicles, and so on. The present invention does not make specific limitations on this.

[0104] Second, the intelligent key of the present invention should also be understood in a broad sense, which refers to a key integrated with advanced technologies, and can be physical (for example, a physical key with physical buttons and / or a touch screen), or virtual (for example, a virtual key implemented by using a mobile phone APP, or a virtual key based on cloud services and capable of being managed and shared through the Internet). The present invention does not make specific limitations on this.

[0105] According to the fourth aspect of the present invention, there is also provided an intelligent entry and start system, including an unknown node, a data processing unit and at least three anchors. Among them, the unknown node is used to be set on a mobile unlocking terminal matched with the vehicle and is used to send positioning signals; the anchor is used to be set on the vehicle and is used to receive the positioning signals sent by the unknown node; the data processing unit is used to be set on the vehicle and is used to execute the positioning method of any one of the technical solutions in the first aspect of the present invention.

[0106] In this way, the intelligent entry and start system of the present invention can achieve fast and accurate positioning between the vehicle and the mobile unlocking terminal in a variety of application environments, and then facilitate the realization of other functions, such as keyless entry, or keyless start, etc.

[0107] The above is only the specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions within the technical scope disclosed by the present invention should be covered by the protection scope of the present invention. Therefore, the protection scope of the present invention should be subject to the protection scope of the claims.

Claims

1. A positioning method, improved based on the trilateral positioning method, characterized in that: The steps include: Step S1: The three anchor points obtain the positioning signal sent by the unknown node, calculate the corresponding RSSI values ​​respectively, and calculate the distances between the three anchor points and the unknown node respectively according to the corresponding RSSI values; Step S2: Draw circles with the three anchor points as the center and the distances between the three anchor points and the unknown node as the radius to obtain three circles, and determine the current states of the three circles according to the positional relationship of the three circles; Step S3: Perform weighted compensation on the three circles according to their current states to obtain the positions of the unknown nodes.

2. The positioning method according to claim 1, characterized in that: In step S2, the current three-circle state includes that the three circles have a common intersection, the three circles intersect in a common area, the three circles intersect in pairs but have no common area, two circles do not intersect, one circle does not intersect with the other two circles, and the three circles are separated; The step S3 specifically includes: For three circles with common intersection points, the coordinates of unknown nodes are calculated using the least squares method; For the common area where three circles intersect, the centroid of the common area is taken as the unknown node; For three circles that intersect each other but have no common area, the centroid point in the middle of the three circles that does not occupy any area is taken as an unknown node; For two circles that do not intersect, a weighted compensation method is used to increase the radius of the two separated circles so that the two separated circles can be tangent after compensation. The compensation factor is the ratio of the distance between the centers of the two separated circles to the sum of the radii. Then, the centroid point of the unoccupied area among the three circles after weighted compensation is taken as the unknown node. If a circle does not intersect with the other two circles, the radius of the circle is increased by weighted compensation so that the circle can intersect with the other two circles. The compensation factor is the largest value of the ratio of the center distance and radius of the circle and the other two circles. Then, the centroid point of the area not occupied by the three circles after weighted compensation is taken as the unknown node. For three circles that are separated from each other, weighted compensation is used to increase the radii of the three circles so that the three circles intersect each other. The compensation factor is the largest value of the ratio of the distance between the centers of each two circles to the sum of the radii of the two circles. Then, the centroid point of the unoccupied area among the three circles after weighted compensation is taken as the unknown node.

3. The positioning method according to claim 1, characterized in that: The step S1 further includes: setting another anchor point and acquiring a positioning signal sent by an unknown node, calculating a corresponding RSSI value and calculating a distance between the anchor point and the unknown node according to the RSSI value; The step S2 comprises: drawing circles with four anchor points as the center and the distances between the four anchor points and the unknown node as the radius to obtain four circles, and combining any three of the four circles to obtain four different current three-circle states; The step S3 specifically includes: performing weighted compensation according to four different current three-circle states to obtain preliminary coordinates of four unknown nodes, and then setting weights according to the sum of the reciprocals of the radii of the circles with the anchor point as the center in each combination, and performing weighted summation on the four preliminary coordinates using the set weights to obtain the positioning coordinates of the unknown nodes.

4. The positioning method according to any one of claims 1 to 3, characterized in that: The positioning method also includes introducing a Kalman filter algorithm to process the positioning signal, specifically including: Step S4-1-1: Get the position coordinates of the unknown node output in step S3, determine whether it is the first time to obtain the position coordinates of the unknown node, if so, set it as the initial positioning coordinates of the unknown node, if not, proceed to step S4-1-2; Step S4-1-2: perform state prediction and covariance prediction in sequence; Among them, the calculation formula for state prediction is: The formula for calculating the covariance forecast is: In the formula, is the estimated value of the state for this prediction, is the state transfer matrix, is the transposed matrix of the state transfer matrix, is the state estimate obtained from the last update, is the covariance matrix obtained in this update, is the covariance matrix obtained from the last update, is the process noise covariance matrix; Step S4-1-3: Calculate Kalman gain, state update and covariance update in sequence; The calculation formula of Kalman gain is: The formula for status update is: The formula for covariance update is: In the formula, is the Kalman gain, is the observation matrix, is the observation matrix The transposed matrix of is the measurement noise covariance matrix, is the estimated value of the state at time k, is the measured value, is the state covariance matrix at time k, is the identity matrix; Step S4-1-4: Use the updated state value as the final position coordinate of the unknown node.

5. The positioning method according to any one of claims 1 to 3, characterized in that: The positioning method also includes introducing an extended Kalman filter algorithm to process the positioning signal, specifically including: Step S4-2-1: Get the position coordinates of the unknown node output in step S3, determine whether it is the first time to obtain the position coordinates of the unknown node, if so, set it as the initial positioning coordinates of the unknown node, if not, proceed to step S4-2-2; Step S4-2-2: perform state prediction and covariance prediction in sequence; Among them, the calculation formula for state prediction is: The formula for calculating the covariance forecast is: In the formula, is the estimated value of the state for this prediction, is the nonlinear state transfer function, is the state estimate obtained from the last update, is the covariance matrix obtained in this update, is the Jacobian matrix of the nonlinear state transfer function, is the covariance matrix obtained from the last update, is the process noise covariance matrix; Step S4-2-3: measuring residuals, calculating residual covariance, calculating Kalman gain, state update and covariance update in sequence; The calculation formula of the measurement residual is: The formula for calculating the residual covariance is: The calculation formula of Kalman gain is: The formula for status update is: The formula for covariance update is: In the formula, is the difference between the measured value and the predicted value, is the measured value, is the predicted value, is the residual covariance, is the Jacobian matrix of the measurement function, yes The transposed matrix of is the uncertainty of the measurement noise, is the Kalman gain, is the estimated value of the state at time k, is the state covariance matrix at time k, is the identity matrix; Step S4-2-4: Use the updated state value as the final position coordinate of the unknown node.

6. A positioning system, characterized in that: The positioning method applied to any one of claims 1 to 5, wherein the positioning system comprises: Unknown node, used to send positioning signals; At least three anchor points, each of which is used to receive a positioning signal sent by the unknown node; The data processing unit is used to calculate the corresponding RSSI values ​​according to the positioning signals, and to calculate the distances between the three anchor points and the unknown nodes according to the corresponding RSSI values, and to determine the current three-circle states, and to perform weighted compensation on the three-circle states according to the current three-circle states to obtain the position of the unknown node.

7. A vehicle, characterized in that: Comprising the positioning system as claimed in claim 6, the anchor point and the data processing unit are arranged on the vehicle.

8. The vehicle according to claim 7, characterized in that The vehicle further includes a smart key, and the unknown node is arranged on the smart key.

9. An intelligent entry and start system, characterized in that: include: Unknown node, used to be set in a mobile unlocking terminal matched with the vehicle and used to send a positioning signal; At least three anchor points, which are arranged on the vehicle and used to receive the positioning signal sent by the unknown node; A data processing unit is used to be arranged on a vehicle and to execute the positioning method as described in any one of claims 1-5.

Images