Golf ball positioning system and method based on bluetooth AOA technology

Abstract

The application relates to a golf positioning system and method based on a Bluetooth AOA technology. According to the erection of a Bluetooth antenna matrix, the in-phase samples and the quadrature samples emitted by a BLE are captured through an antenna array, the phase difference of the radio signals received by each element of the antenna array is calculated through the samples, a series of phase differences are taken as parameters, the spatial fitting algorithm is used to calculate the golf position coordinates, the linear regression algorithm is used to calculate the motion line according to the dynamic positioning trajectory of the golf, the positioning coordinates are finally predicted, and the midpoint of the two coordinates is taken as the final coordinate. The positioning error correction algorithm and the linear regression algorithm are used to establish different dynamic prediction motion trajectories for different terrains and different weathers, the feature prediction value is adjusted according to the prediction error of the current area, the positioning accuracy is high, the hardware device of the Bluetooth BLE5.1 or above version is used, the working range is wide, the network point quantity is small, the cost is low, and the energy consumption is low.

Description

Technical Field

[0001] This invention belongs to the field of positioning technology and relates to a golf ball positioning system and method based on Bluetooth AOA technology. Background Technology

[0002] Traditional positioning technologies mainly include satellite positioning, ultrasonic positioning, base station positioning, RFID positioning, and Bluetooth RSSI positioning. However, all of these technologies have various problems, such as complex implementation methods, high costs, and low accuracy. For golf positioning systems, it is necessary to accurately locate the landing point of the golf ball to reduce the time spent searching for the ball.

[0003] The Bluetooth SIG has released the Bluetooth BLE 5.1 ​​standard, which features direction-finding capabilities and positioning accuracy down to the centimeter level, while Bluetooth RSSI offers positioning accuracy from 1 to 10 meters. It provides two different positioning methods: Angle of Arrival (AoA) and Angle of Departure (AOD). The basic principle is that the transmitter emits a sine wave signal, and the receiver receives the signal and calculates the phase. Therefore, a golf ball positioning system and method based on Bluetooth AOA technology is proposed for the precise location of the golf ball's landing point, facilitating ball retrieval. Summary of the Invention

[0004] The purpose of this invention is to provide a golf ball positioning system and method based on Bluetooth AOA technology. The system involves setting up a Bluetooth antenna matrix, capturing in-phase and quadrature samples emitted by Bluetooth through an antenna array, and calculating the phase difference of the radio signals received by the antenna array using the sample data. Then, using a series of phase differences as parameters, the golf ball's position coordinates are calculated according to a spatial fitting algorithm. Next, the dynamic positioning trajectory of the golf ball over a time period is used to calculate the movement path using a linear regression algorithm to predict the ball's final coordinates. Finally, the coordinates of the final coordinate point are optimized again using influencing factors.

[0005] This invention is based on a Bluetooth beacon linear matrix and a Bluetooth chip. The Bluetooth beacon linear matrix is ​​uniformly deployed within the golf course, and Bluetooth antenna elements are placed along a rectangular grid in a 4x4 array. The course is divided into 10m x 10m zones, and an influencing factor database is established for each zone, storing influencing factors for different weather and terrain conditions to correct prediction errors. A Bluetooth chip supporting Bluetooth BLE 5.1 ​​is installed inside the golf ball to transmit position signals.

[0006] Based on the above system, the following methods for positioning golf balls are proposed:

[0007] When a golf ball lands in a rectangular antenna matrix, the four points in the matrix are matched in pairs. The in-phase samples and orthogonal samples are calculated by the phase difference of different antennas, and then the six sets of calculation results are sent to the cloud positioning server.

[0008] After receiving the data, the positioning server calculates six sets of actual coordinates using the AoA algorithm, and then uses a positioning error correction algorithm to calculate the precise position Pos at time t. t .

[0009] Before the golf ball stops moving, the positioning server calculates the position value of the golf ball with a set period N, and fits the trajectory curve function LinearRegPos of the golf ball within the period N using a linear regression algorithm.

[0010] Once the golf ball stops moving, the coordinates Pos0 are first calculated using LinearRegPos. Then, based on Pos0 and the current weather conditions, the predicted feature value σ for the area where Pos0 is located is retrieved from the influence factor database. his Calculate the coordinates Pos1, and use the midpoint between these two coordinates as the final coordinates.

[0011] After the golf ball is located, its actual coordinates Pos are sent via manual measurement. r Calculate the prediction error σ for the region. cur When the error in this region reaches a set threshold, the feature prediction value σ of this region in the impact factor database is adjusted. his .

[0012] Compared with the prior art, the beneficial effects of the present invention are:

[0013] Precise positioning: It adopts a spatial fitting error correction algorithm (positioning error correction algorithm) based on angular positioning technology, which reduces the traditional positioning error from 1-10 meters to 20-40 cm. Since the diameter of a golf ball is generally 42.67 mm, the error of 20-40 cm is more suitable for finding the ball than 1-10 meters.

[0014] Dynamic prediction: By using a linear regression algorithm, the final coordinates are predicted not only from a static, single signal, but also through continuous points in motion and curved trajectories, significantly improving accuracy. Different linear regression prediction values ​​are established for different terrains and weather conditions, accurately describing the impact of terrain and weather on the movement trajectory. This fully considers the influence of weather and terrain on positioning accuracy, improving computational precision.

[0015] Artificial intelligence fine-tuning: The prediction error of the current area is calculated using real coordinates, and the feature prediction values ​​are adjusted according to the prediction error of the current area to further improve the accuracy of positioning;

[0016] Wide range and low cost: Using hardware devices with Bluetooth BLE 5.1 ​​or higher, the theoretical range can reach 300 meters, with an actual range of about 100 meters. The wide working range reduces the number of network deployment points and lowers costs.

[0017] Low power consumption and energy saving: Through the Bluetooth BLE protocol, the amount of data exchanged each time is small and the connection is fast, with a communication time of 2-3 milliseconds. The actual working time for 20,000 connections is only 1 minute. When not connected, it is in a sleep state with low power consumption and can work for up to 1 year after a single charge. Attached Figure Description

[0018] Figure 1 Bluetooth beacon deployment map for a golf course;

[0019] Figure 2 This is a table of influence factors for a coordinate point in one of the two regions in the example;

[0020] Figure 3 A structural diagram of a Bluetooth chip and antenna built into a golf ball;

[0021] Figure 4 Logic diagram for Bluetooth BLE 5.1 ​​AoA angle positioning;

[0022] Figure 5 Generate a logic diagram with 6 coordinates for 4 locators;

[0023] Figure 6 The geometry for the localization error correction algorithm;

[0024] Figure 7 This is a coordinate graph of linear regression.

[0025] Figure 8 To correct the location map using the predicted values ​​of the impact factors;

[0026] Figure 9 This is a logic diagram of the prediction process. Detailed Implementation

[0027] The following detailed description, in conjunction with the accompanying drawings, outlines the specific implementation steps of the golf ball positioning system and method based on Bluetooth AOA technology of the present invention.

[0028] Establish a Bluetooth locator matrix in the golf course, such as Figure 1 As shown, the matrix is ​​arranged in a square with a spacing of N (N is generally between 50 and 100 meters, and this embodiment takes 100 meters as an example) and a height of H above the ground (H is generally between 40 and 100 centimeters).

[0029] Establish such a database Figure 2 The diagram shows the influencing factors for different regions and weather conditions. Weather is categorized into four types: sunny, cloudy, rainy, and snowy. The region type is obtained through coordinates. During system initialization, the influencing factors of the sphere's movement from rolling to stopping under each of the four weather conditions are measured, i.e., the characteristic prediction values ​​of linear programming. Each weather condition is sampled more than 100 times, and the average value is calculated.

[0030] like Figure 3 As shown, a chip supporting Bluetooth version 5.1 or higher is installed inside the golf ball. Considering the golf ball's size of 42.67mm, the chip's length and width must both be no greater than 20mm for practical feasibility. This embodiment selects a Nordic E73-2G4M08S1C chip. A spherical metal antenna mesh (connected to the chip via two antennas) is wrapped around the inner surface of the ball's outer surface to transmit wireless signals. Finally, a wear-resistant rubber layer is wrapped around the outside to ensure the ball's strength. To ensure the ball's airtightness, a lithium battery coil module is installed inside the golf ball to power the Bluetooth chip, and the lithium battery coil module is charged wirelessly.

[0031] like Figure 4 As shown, when the golf ball stops rolling and becomes fixed inside a beacon rectangle, the Bluetooth chip inside transmits a CTE (Constant Tone Extension) signal to surrounding Bluetooth beacons via its antenna. When the locator's antenna array receives the signal, due to the different phase differences between the antennas, it uses the Bluetooth Angle of Arrival (AOA) algorithm, based on the signal wavelength, the distance between the antennas, and the phase of the received signal, to send the angle of each antenna and the sample dataset IQ to the positioning server.

[0032] like Figure 5 As shown, after the positioning server receives the angle data, it can predict the position coordinates (P1 to P6) of six golf balls by using different combinations of four rectangular locators (ApB / ApC / BpD / CpD / ApD / BpC) based on the angles of the two antennas and the sample dataset.

[0033] like Figure 6 As shown, according to the positioning error correction algorithm, the final positioning coordinates are calculated using 6 coordinate positions. Specifically, first, the nearest adjacent points among the 6 points are connected to form a hexagon. Then, the diagonal points of the hexagon are connected with straight lines, and the 3 diagonals form a triangle inside. Then, a ray is drawn at the bisection point of each interior angle of the triangle. The intersection of the 3 rays is the final positioning coordinate.

[0034] Given the coordinates of four points P0(x0,y0), P1(x1,y1), P2(x2,y2), and P3(x3,y3), calculate the coordinates (x,y) of the intersection point of any two diagonals using the following formula:

[0035] The coordinates of the three intersection points Q1, Q2, and Q3 are calculated using the above formula. The incenter of the triangle is then calculated, ultimately yielding the precise coordinates of the golf ball's location (Z). x Zy ):

[0036] Where D1 represents the distance between Q1 and Q2, D2 represents the distance between Q2 and Q3, and D3 represents the distance between Q1 and Q3; Q 1.X Q represents the x-axis coordinate of Q1, Q 1.y This represents the y-axis coordinate of Q1; Q 2.X Q represents the x-axis coordinate of Q2, Q 2.y This represents the y-axis coordinate of Q2; Q 3.X Q represents the x-axis coordinate of Q3, Q 3.y This represents the y-axis coordinate of Q3.

[0037] Taking P0 and P1 as an example, the distance between two points is calculated using the distance function f(﹒). The specific expression for f(﹒) is as follows:

[0038]

[0039] According to the above formula, we get: D1 = f(Q1,Q2), D2 = f(Q2,Q3), D3 = f(Q1,Q3).

[0040] like Figure 7 As shown, the server continuously monitors the signal within each area. When a golf ball lands in a certain area, the server calculates its position value Pos(t) at time t every 100 milliseconds until it stops rolling. Then, using the continuous position data of the golf ball for N seconds (default value 5) before it stops moving, a univariate linear regression algorithm is used to fit its trajectory curve function LinearRegPos and the feature prediction value σ of this linear regression function. cur .

[0041] like Figure 8 As shown, after the golf ball stops moving, the coordinates Pos0 are first calculated based on LinearRegPos. Then, based on Pos0 and the current weather conditions, the characteristic prediction value σ of the area where Pos0 is located is retrieved from the influence factor database. his Correct the final coordinate position Pos f , Pos 0.x Pos represents the x-axis coordinate of Pos0. 0.y This represents the y-axis coordinate of Pos0.

[0042] Once the golf ball is located, its true coordinates (Pos) are sent via manual testing. r Calculate the prediction error σ for the current region. curWhen the number of prediction error samples in a certain region exceeds a certain threshold (e.g., 1000 times), or after the system has been running for a certain period of time (e.g., one month), the feature prediction values ​​σ of the impact factor database are adjusted using the least squares algorithm. his This further improves the accuracy of positioning.

[0043] like Figure 9 As shown, the above feature prediction optimization process is as follows: First, the position information Pos(t) at each moment is continuously calculated before the sphere stops; after the sphere stops, the position coordinates set before stopping N seconds are fitted to the motion trajectory curve using a linear regression algorithm; then, based on the feature prediction values ​​in the impact factor database, the final position is optimized by calculating the error; finally, the error value is calculated based on the actual position. When the error data in a certain area reaches a certain amount (threshold), the impact factor database is updated using the least squares algorithm.

Claims

1. A golf ball positioning method based on Bluetooth AOA technology, characterized in that: The system is implemented as follows: It includes a Bluetooth beacon linear matrix and a Bluetooth chip. The Bluetooth beacon linear matrix is ​​evenly distributed throughout the golf course, with Bluetooth antenna elements placed along a rectangular grid in a 4x4 array. A database of influencing factors is established within 10m x 10m zones of the course, storing influencing factors for different weather and terrain conditions in each zone to correct prediction errors. A Bluetooth chip supporting Bluetooth BLE 5.1 ​​is installed inside each golf ball to transmit position signals. The method specifically includes the following steps: When a golf ball lands in a rectangular antenna matrix, the four points in the matrix are matched in pairs. The in-phase samples and orthogonal samples are calculated by the phase difference of different antennas, and then the six sets of calculation results are sent to the cloud positioning server. After receiving the data, the positioning server calculates six sets of actual coordinates using the AoA algorithm, and then uses a positioning error correction algorithm to calculate the precise position Pos at time t. t ; The positioning and error correction algorithm is as follows: First, connect the nearest adjacent points among the 6 points to form a hexagon. Then, connect the diagonal points of the hexagon with straight lines. The 3 diagonals form a triangle inside. Then, draw a ray at the bisection point of each interior angle of the triangle. The intersection of the 3 rays is the final positioning coordinate. Before the golf ball stops moving, the positioning server calculates the position value of the golf ball with a set period N, and fits the trajectory curve function LinearRegPos of the golf ball within the period N using a linear regression algorithm; Once the golf ball stops moving, the coordinates Pos0 are first calculated using LinearRegPos. Then, based on Pos0 and the current weather conditions, the predicted feature values ​​for the area where Pos0 is located are retrieved from the impact factor database. Calculate the coordinates Pos1, and use the midpoint between these two coordinates as the final coordinates; After the golf ball is located, its actual coordinates Pos are sent via manual measurement. r Calculate the prediction error for the region. cur When the error in this region reaches a set threshold, the least squares algorithm is used to adjust the feature prediction value of this region in the impact factor database. .

2. The golf ball positioning method as described in claim 1, characterized in that: The golf ball contains a lithium battery coil module that powers the Bluetooth chip, and the lithium battery coil module is charged wirelessly.

3. The golf ball positioning method as described in claim 1, characterized in that: The length and width of the chip are both no greater than 20mm.

Images