Location estimation system and location estimation method

Abstract

To provide a position estimation system and a position estimation method capable of more accurately estimating a position of a portable device on the basis of a signal transmitted from the portable device, even if a data amount is small.SOLUTION: According to a position estimation system and a position determination method, reception intensity and a ToF related value in each of a plurality of anchors 30 are used in order to estimate positions of portable devices 103. The ToF related value is corrected by using a correction function that outputs a corrected ToF related value by inputting the acquired reception intensity and ToF related value. The correction function is a function determined using teacher data including the positions of the plurality of portable devices 103, and the reception intensity and ToF related value at the positions of the plurality of portable devices 103. The correction function is generated such that a data amount is smaller than that of the teacher data. This can reduce the data amount more than when saving the all pieces of teacher data and correcting using the teacher data.SELECTED DRAWING: Figure 3

Description

[Technical Field]

[0001] The present disclosure relates to a position estimation system and a position estimation method that are mounted on a vehicle and used to estimate the relative position of a portable device carried by a user with respect to the vehicle based on the reception conditions of wireless signals transmitted from the portable device. [Background technology]

[0002] Conventionally, electronic key systems are known that allow or execute certain vehicle operations, such as locking and unlocking the vehicle and starting the engine, based on wireless communication between an in-vehicle device and an electronic key. Such electronic key systems allow or execute certain operations depending on the location of the electronic key relative to the vehicle. Therefore, it is necessary to estimate the location of the electronic key.

[0003] One method for determining the location of an electronic key relative to a vehicle is to use the strength of the vehicle signal received by the electronic key. Specifically, the location of the electronic key is determined based on a combination of the strength of the vehicle signal received by the electronic key from one vehicle antenna and the strength of the vehicle signal received by the electronic key from another vehicle antenna. This is because the strength of the received signal is correlated with distance, as it weakens as the distance from the transmitter increases.

[0004] One example of a method for improving the accuracy of estimating the location of an electronic key is disclosed in Patent Document 1. The electronic key system described in Patent Document 1 takes into account individual differences in the reception sensitivity of electronic keys and estimates the location of the electronic key by correcting the received signal strength using the correlation between the reception sensitivity of the electronic key and the received signal strength.

[0005] Another method for determining the location of an electronic key is a distance measurement method that uses the time of flight (ToF) of a radio signal. [Prior art documents] [Patent documents]

[0006] [Patent Document 1] Japanese Patent Application Publication No. 2019-85734 Summary of the Invention [Problem to be solved by the invention]

[0007] When using a ToF distance measurement method to estimate the location of an electronic key, if there is an obstacle such as the vehicle body or a human body between the antenna mounted on the vehicle and the electronic key, the measurement distance will be extended due to diffraction of radio waves or the shadow of the obstacle, etc., which means that direct waves cannot be detected and only reflected waves will be detected, resulting in a large error in the location estimation.

[0008] Furthermore, when estimating the location of an electronic key, the accuracy of the location estimation can be improved by using actual measurement data acquired in advance and comparing the actual measurement data with the measurement data. The more measurement points there are for the actual measurement data, the more accurate the location estimation will be, but there is also the problem of the amount of data required.

[0009] Therefore, the disclosed object has been made in consideration of the above-mentioned problems, and aims to provide a position estimation system and a position estimation method that can more accurately estimate the position of a mobile device based on a signal emitted from the mobile device, even with a small amount of data. [Means for solving the problem]

[0010] The present disclosure employs the following technical means to achieve the above-mentioned objectives.

[0011] The position estimation system disclosed herein is a position estimation system that estimates the position of a portable device (103) carried by a user relative to a vehicle (101) by wirelessly communicating with the portable device, and includes: a plurality of communication devices (30) configured to be able to wirelessly communicate with the portable device and installed at different positions on the vehicle; a communication control unit (20) that controls the operation of the plurality of communication devices; and a position estimation unit (20) that estimates the position of the portable device, wherein the communication control unit acquires the reception strength of a signal from the portable device at each of the plurality of communication devices, and acquires a flight time-related value that is a parameter separate from the reception strength and that directly or indirectly indicates the flight time of radio waves from each of the plurality of communication devices to the portable device, and the position estimation unit corrects the flight time-related value using a correction function that inputs the acquired reception strength and flight time-related value and outputs a corrected flight time-related value, and estimates the position of the portable device using the corrected flight time-related value, and the correction function is a function determined using training data including the positions of the plurality of portable devices and the reception strength and flight time-related values ​​at the positions of the plurality of portable devices, and the position estimation system has a smaller amount of data than the training data.

[0012] Furthermore, the position estimation method disclosed herein is a position estimation method executed by at least one processor to estimate the position of a portable device carried by a user, and includes the steps of: acquiring the reception strength of a signal from the portable device from a plurality of communication devices configured to be able to wirelessly communicate with the portable device and each being located at a different position in the vehicle; acquiring a flight time-related value which is a parameter separate from the reception strength and which directly or indirectly indicates the flight time of radio waves from each of the plurality of communication devices to the portable device; correcting the flight time-related value using a correction function which inputs the acquired reception strength and flight time-related value and outputs a corrected flight time-related value; and estimating the position of the portable device using the corrected flight time-related value, wherein the correction function is a function determined using training data including the positions of the plurality of portable devices and the reception strength and flight time-related values ​​at the positions of the plurality of portable devices, and this is a position estimation method with a smaller amount of data than the training data.

[0013] According to such a position estimation system and position determination method, the reception strength and time-of-flight-related value of each of multiple communication devices are used to estimate the position of a mobile device. The reception strength and the time-of-flight-related value are different parameters, and the factors that cause errors may differ. Therefore, the estimated position of the mobile device when using only the reception strength may not match the estimated position of the mobile device when using only the time-of-flight-related value. Therefore, the time-of-flight-related value is corrected using a correction function that inputs the acquired reception strength and time-of-flight-related value and outputs a corrected time-of-flight-related value.

[0014] The correction function is a function determined using teacher data including the positions of multiple mobile devices and the reception strength and time-of-flight related values ​​at the positions of the multiple mobile devices. The correction function is generated so that the data volume is smaller than that of the teacher data. This allows the data volume to be smaller than if all of the teacher data were stored and correction were performed using the teacher data. Furthermore, since the time-of-flight related values ​​are corrected using a correction function that uses two parameters, the correction accuracy can be improved compared to a configuration in which correction is performed using only one parameter. As a result, the corrected time-of-flight related values ​​can be used to more accurately estimate the position of the mobile device.

[0015] The symbols in parentheses for the above-mentioned means are examples showing the correspondence with the specific means described in the embodiments to be described later. [Brief explanation of the drawings]

[0016] [Figure 1] FIG. 1 is a block diagram showing the configuration of an electronic key system. [Figure 2] FIG. 10 is a diagram showing an example of anchor placement. [Figure 3] 1 is a flowchart illustrating a location estimation method. [Figure 4] FIG. 1 is a diagram for explaining a position estimation method. [Figure 5] A diagram explaining a neural network. DETAILED DESCRIPTION OF THE INVENTION

[0017] (First embodiment) A first embodiment of the present disclosure will be described with reference to FIGS. 1 to 5. An electronic key system 100 of this embodiment is applied to a vehicle 101. As shown in FIG. 1, the electronic key system 100 includes an in-vehicle system 102 and a portable device 103. The in-vehicle system 102 is a system installed in the vehicle 101. The portable device 103 is a device carried by the user of the vehicle 101. There may be multiple portable devices 103. The electronic key system 100 corresponds to a position estimation system, and estimates the position of the portable device 103 relative to the vehicle 101 by wirelessly communicating with the portable device 103.

[0018] First, the underlying configuration of the electronic key system 100 will be described. The in-vehicle system 102 and the portable device 103 are configured to be capable of short-range communication with each other. Here, short-range communication refers to communication that complies with a predetermined short-range wireless communication standard, with an effective communication distance of, for example, approximately 5 to 100 meters. In this embodiment, the short-range communication standards used are, for example, Bluetooth Low Energy (BLE; Bluetooth is a registered trademark) and Ultra Wide Band Impulse Radio (UWB-IR).

[0019] The in-vehicle system 102 and the portable device 103 are each configured to be able to perform wireless communication in accordance with the BLE standard. Therefore, the details of the communication method, such as communication connection and encrypted communication, are performed according to the sequence defined by the BLE standard.

[0020] The following describes a case where the in-vehicle system 102 is configured to act as a master in communication with the mobile device 103, and the mobile device 103 is configured to act as a slave. In BLE communication, a slave is a device that intermittently transmits advertising signals and transmits and receives data based on requests from the master. The master is a device that controls the communication connection state and communication timing with the slave.

[0021] An advertising signal is a signal for notifying other devices of its own existence. Signals transmitted and received using BLE, such as advertising signals, include sender information. The sender information is, for example, unique identification information assigned to the mobile device 103. For example, a device address or a universally unique identifier (UUID) can be used as the device identification information.

[0022] The portable device 103 is a device that holds key information for using the vehicle 101 and functions as an electronic key for the vehicle 101 using the key information. The key information here is data used in an authentication process, which will be described later. The key information is data for verifying that a person attempting to access the vehicle 101 is a legitimate user, that is, for verifying the legitimacy of the person attempting to access the vehicle 101. The key information can be called an authentication key, an encryption key, or a key code. The key information may differ for each portable device 103. The in-vehicle system 102 stores and registers the key information for each portable device 103 in association with device identification information.

[0023] The in-vehicle system 102 performs automatic authentication processing via wireless communication with the portable device 103. Then, on the condition that the authentication is successful, a system is realized that performs vehicle control according to the user's position relative to the vehicle 101. The vehicle control here includes locking and unlocking the doors, turning the power on and off, starting the engine, etc.

[0024] For example, when the in-vehicle system 102 confirms that the portable device 103 is within a locking / unlocking area preset for the vehicle 101, the in-vehicle system 102 executes control such as locking or unlocking the doors based on a user operation on a door button. The locking / unlocking area is set outside the vehicle cabin near the driver's door, the passenger door, and the trunk door.

[0025] Furthermore, when the in-vehicle system 102 has confirmed through wireless communication with the portable device 103 that the portable device 103 is present in the vehicle, the in-vehicle system 102 executes engine start control based on a user operation on the start button.

[0026] The authentication of the portable device 103 by the in-vehicle system 102 may be performed, for example, by a challenge-response method. The authentication process is a process of comparing a response code generated by the portable device 103 based on key information with a verification code held by or dynamically generated by the vehicle 101.

[0027] Next, the portable device 103 will be described. The portable device is a portable, general-purpose information processing terminal equipped with a communication function. A digital key app, which is an application for functioning as an electronic key for the vehicle 101, is installed on the portable device 103. For example, a smartphone, a tablet terminal, a wearable device, or the like can be used as the portable device 103. A wearable device is a device that is worn on the user's body and can be in a variety of shapes, such as a wristband type, a watch type, a ring type, glasses type, or earphone type.

[0028] 1, the portable device 103 includes a terminal control unit 10, a portable BLE module 11, and a portable UWB module 12. In this embodiment, for convenience, a description of the configuration of the portable device 103 other than that related to wireless communication with the in-vehicle system 102 is omitted.

[0029] The portable BLE module 11 is a communication module capable of performing short-range wireless communication conforming to the aforementioned BLE standard. The portable BLE module 11 is configured, for example, with an IC, an antenna, a communication circuit, etc. The portable BLE module 11 establishes a communication connection with the in-vehicle system 102 to perform short-range wireless communication.

[0030] The portable BLE module 11 periodically scans and receives advertising signals periodically transmitted from the in-vehicle system 102. Upon receiving the advertising signal, the portable BLE module 11 transmits a connection request to the in-vehicle system 102. If this connection request is accepted, a communication connection between the portable BLE module 11 and the in-vehicle system 102 is established.

[0031] The portable UWB module 12 is a communication module capable of performing short-range wireless communication using the UWB-IR system. Hereinafter, short-range wireless communication using the UWB-IR system will be referred to as UWB communication. UWB communication can also be referred to as ultra-wideband wireless communication. Like the portable BLE module 11, the portable UWB module 12 is configured with, for example, an IC, an antenna, a communication circuit, and the like.

[0032] Portable UWB module 12 performs UWB communication by transmitting and receiving impulse-shaped radio waves (hereinafter, impulse signals). An impulse signal used in UWB communication is a signal with an extremely short pulse width. For example, the pulse width is a signal of 2 ns. Furthermore, an impulse signal used in UWB communication is a signal with a bandwidth of 500 MHz or more (i.e., an ultra-wide bandwidth). Frequency bands available for UWB communication (hereinafter, UWB bands) include 3.1 GHz to 10.6 GHz, 3.4 GHz to 4.8 GHz, 7.25 GHz to 10.6 GHz, and 22 GHz to 29 GHz. When portable UWB module 12 receives an impulse signal transmitted from in-vehicle system 102, it replies with a response signal corresponding to the signal.

[0033] The terminal control unit 10 is realized by a central processing unit (abbreviated as CPU) and controls the portable BLE module 11 and the portable UWB module 12. The terminal control unit 10 executes processes for realizing the functions of each unit of the portable device 103. The terminal control unit 10 executes programs stored in memory, thereby enabling communication with the in-vehicle system 102.

[0034] Next, the in-vehicle system 102 will be described. As shown in Fig. 1, the in-vehicle system 102 includes a smart ECU 20 and multiple anchors 30. ECU stands for Electronic Control Unit, and refers to an electronic control device. The anchors 30 correspond to communication devices.

[0035] The smart ECU 20 is connected to each of the multiple anchors 30 via an in-vehicle network 104. The smart ECU 20 is also connected to other in-vehicle control devices, such as a power supply ECU and a body ECU, via the in-vehicle network 104 so as to be able to communicate with each other. The in-vehicle network 104 is a communication network established within the vehicle. The in-vehicle network 104 is based on a standard such as Controller Area Network (CAN: registered trademark).

[0036] The smart ECU 20 estimates the position of the portable device 103 in cooperation with the anchor 30 and the like. The smart ECU 20 also performs vehicle control in accordance with the result of estimating the position of the portable device 103 in cooperation with other ECUs. The smart ECU 20 is implemented using a computer. That is, the smart ECU 20 includes a processor 21, RAM, storage, I / O, and bus lines connecting these components. Of these components, only the processor 21 is shown in FIG. 1.

[0037] The processor 21 is hardware for arithmetic processing coupled with RAM (Random Access Memory). The processor 21 is, for example, a CPU. The processor 21 accesses the RAM to execute various processes for realizing the functions of each functional unit. The RAM is a volatile storage medium. The storage includes a non-volatile storage medium such as a flash memory. The storage stores various programs to be executed by the processor 21. The execution of a program by the processor 21 corresponds to the execution of a method corresponding to the program, for example, a position estimation method. The I / O is a circuit module for communicating with other devices. The I / O is realized using analog circuit elements, ICs, etc.

[0038] The storage registers device identification information for each portable device 103. The storage also stores setting data indicating the mounting position of each anchor 30 in the vehicle 101. The storage corresponds to a storage device in which the setting data is stored.

[0039] The smart ECU 20 includes an in-vehicle BLE module 22 and an in-vehicle UWB module 23. The in-vehicle BLE module 22 is a communication module capable of performing short-range wireless communication in accordance with the above-mentioned BLE standard. The in-vehicle BLE module 22 has the same configuration as the above-mentioned portable BLE module 11. The in-vehicle BLE module 22 establishes a communication connection with the portable device 103 to perform short-range wireless communication.

[0040] The in-vehicle UWB module 23 is a communication module capable of the above-mentioned UWB communication. The in-vehicle UWB module 23 has the same configuration as the above-mentioned portable UWB module 12. The in-vehicle UWB module 23 performs UWB communication by transmitting and receiving impulse signals. When the in-vehicle UWB module 23 receives an impulse signal transmitted from the portable device 103, it returns a response signal corresponding to the signal.

[0041] The multiple anchors 30 are configured to be capable of wireless communication with the mobile device 103 and are installed at different positions in the vehicle 101. The anchors 30 are communication modules including an anchor UWB module 31 and an anchor control unit 32. The anchors 30 perform UWB communication according to instructions from the smart ECU 20. The anchors 30 correspond to antennas. The anchors 30 are installed at multiple locations inside and outside the vehicle 101. Two or more anchors 30 are installed in the vehicle 101, including an anchor 30 installed inside the vehicle cabin and an anchor 30 installed outside the vehicle cabin. Each anchor 30 is set with a unique identifier.

[0042] An example of the arrangement of the anchors 30 in this embodiment will now be described with reference to Fig. 2. As shown in Fig. 2, a total of seven anchors 30 are arranged, four outside the vehicle cabin and three inside the vehicle cabin. For ease of understanding, Fig. 1 shows only three anchors 30. Furthermore, the configurations of the multiple anchors 30 are the same.

[0043] The four anchors 30 outside the vehicle cabin are specifically located near the left and right corners of the front end and the left and right corners of the rear end of the vehicle 101. The three anchors 30 inside the vehicle cabin are specifically located one inside the instrument panel in the front center of the vehicle cabin and one on each side of the rear seat.

[0044] The anchor control unit 32 acquires a time-of-flight related value by transmitting and receiving an impulse signal. The time-of-flight related value is also called a ToF related value. The ToF related value is a parameter that directly or indirectly indicates the flight time of a radio wave from the anchor 30 to the portable device 103. The distance from the anchor 30 to the portable device 103 corresponds to the time of flight (ToF) of the signal.

[0045] ToF is determined based on the two-frequency phase difference and round-trip time (RTT). The two-frequency phase difference and RTT correspond to ToF-related values. ToF-related values ​​can also be called distance-related values. These ToF-related values ​​are parameters different from reception strength. The two-frequency phase difference here is the difference between the transmit and receive phase differences observed at two different frequencies. The two-frequency phase difference corresponds to the amount of phase angle displacement due to frequency changes. The transmit and receive phase difference corresponds to the phase difference between the transmitted CW signal and the received CW signal. The transmit and receive phase difference is also simply called the phase angle.

[0046] In this embodiment, the round trip time is used as the ToF-related value. Specifically, the anchor control unit 32 measures the round trip time, which is the elapsed time from transmitting an impulse signal to receiving an impulse signal as a response signal to the impulse signal. The anchor control unit 32 outputs the measured round trip time together with its own identification information to the smart ECU 20. Note that the method for acquiring the round trip time described here is merely an example. Other methods may be used as long as they can acquire the round trip time used for measuring the distance between the anchor 30 and the portable device 103.

[0047] Furthermore, the anchor control unit 32 measures the strength of the signal received by the anchor UWB module 31. The anchor control unit 32 is configured to sequentially detect the strength of the signal received by the anchor UWB module 31. The signal indicating the reception strength or the measurement value itself is also called RSSI (Received Signal Strength Indicator / Indication). The anchor control unit 32 outputs the measured reception strength together with its own identification information to the smart ECU 20.

[0048] Next, a description will be given of the control of the smart ECU 20. The smart ECU 20 causes the in-vehicle BLE module 22 to transmit an advertising signal. For example, when the vehicle 101 is parked and a certain time has elapsed since all the doors of the vehicle 101 were locked, the smart ECU 20 causes the advertising signal to be transmitted periodically.

[0049] The smart ECU 20 acquires information related to wireless communication with the mobile device 103 from the in-vehicle BLE module 22. When a connection is established between the mobile BLE module 11 of the mobile device 103 that has received the advertising signal and the in-vehicle BLE module 22, the smart ECU 20 acquires information received by the in-vehicle BLE module 22.

[0050] The smart ECU 20 also functions as a communication control unit and controls the operations of the multiple anchors 30. The smart ECU 20 causes the in-vehicle UWB module 23 to transmit an impulse signal. Similarly, the smart ECU 20 causes the anchor UWB module 31 of each anchor 30 to transmit an impulse signal.

[0051] The smart ECU 20 causes the impulse signals to be transmitted in order from the plurality of anchors 30. The smart ECU 20 causes the impulse signals to be transmitted in order from each anchor 30, for example, at predetermined time intervals.

[0052] The smart ECU 20 starts transmitting an impulse signal when a connection with the portable BLE module 11 of the portable device 103 is established or is triggered by the establishment of the connection. This makes it possible to reduce wasteful transmission of an impulse signal even when the portable device 103 is not present around the vehicle 101.

[0053] The smart ECU 20 acquires information about UWB communication received by the anchor 30 when UWB communication is performed between the UWB module of the portable device 103 and the anchor 30. The smart ECU 20 acquires the round trip time output from each of the multiple anchors 30 as a ToF-related value. Therefore, the smart ECU 20 acquires the ToF-related value from each of the multiple anchors 30. The smart ECU 20 also acquires the reception strength from each of the multiple anchors 30. Therefore, the smart ECU 20 acquires the reception strength of the signal from the portable device 103 at each of the multiple anchors 30.

[0054] Next, a method for estimating the position of the portable device 103 will be described. The process shown in Fig. 3 is repeatedly executed in a short period of time by the smart ECU 20. The smart ECU 20 also functions as a position estimation unit, and estimates the position where the portable device 103 is located.

[0055] In step S1, the ToF-related value and reception strength are acquired from each anchor 30, and the process proceeds to step S2. In step S2, the ToF-related value is corrected using a correction function, and the process proceeds to step S3. The correction function inputs the reception strength and ToF-related value acquired in step S1 and outputs the corrected ToF-related value.

[0056] In step S3, it is determined whether the portable device 103 is inside or outside the vehicle, and the process proceeds to step S4. To determine whether the portable device 103 is inside the vehicle, the determination may be made using the corrected ToF-related value, the determination may be made using only the reception strength, or the determination may be made using the uncorrected ToF-related value and the reception strength.

[0057] In step S4, a detailed area determination is performed for the interior or exterior of the vehicle determined in step S3, and the process proceeds to step S5. The detailed area determination is a process of determining which area the vehicle is in among the divided areas obtained by further dividing the interior or exterior of the vehicle. The area referred to here may be an area with a certain area, or may be a process of specifying any one point among multiple discrete points.

[0058] In step S5, the position of the portable device 103 is estimated using the corrected ToF-related values, and this flow ends. In step S5, a specific point within the area determined in step S4 is used as a tentative position, and calculation processing is performed to determine a position with higher accuracy. The specific point within the area may be any point, the center of gravity of the area, or any vertex of the area. In step S5, positioning is performed using, for example, the nonlinear least squares method. However, since the nonlinear least squares method is a method that performs repeated calculations from the initial position to find the point with the smallest error, the initial position is an important factor, and if the initial position is incorrect, it will converge to a local solution, leading to erroneous determination.

[0059] Therefore, in step S4, a detailed area determination is performed. The position estimation in steps S4 and S5 will be explained using Fig. 4. For ease of understanding, Fig. 4 explains using ranging circles C1 to C3 of the three anchors 30 located on the right side of the seven anchors 30. The actual position of the portable device 103 is indicated by a triangle.

[0060] The ranging circles C1 to C3 in FIG. 4 are circles whose centers are the anchors 30 and whose radii are the distances determined using the corrected ToF-related values ​​in step S2. Ideally, the ranging circles C1 to C3 should intersect at a single point at the position of the portable device 103. However, as shown in FIG. 4, there are cases where the three ranging circles C1 to C3 do not intersect at a single point. This is due to various measurement errors.

[0061] In step S5, an initial position must be set in order to estimate the position of the portable device 103 by the nonlinear least squares method. Therefore, in step S4, a tentative position where the three ranging circles C1 to C3 do not intersect at a single point but where the portable device 103 is likely to be located is set, and the initial position is set. The tentative position is a point within the detailed area set in step S4.

[0062] In the example shown in FIG. 4, there are six intersections of the three ranging circles C1 to C3. Of these, it is preferable to set the intersection closest to the actual position of the portable device 103, i.e., the position of the triangle in FIG. 4, as the initial position. Therefore, the initial position is set using a statistical model trained using training data, as will be described later. In the example shown in FIG. 4, the point where the two ranging circles C1 and C3 intersect is set as the initial position, as indicated by a diamond. Then, in step S5, this initial position is used as the start point and a mathematical search is conducted to find a position where the sum of the distances from the ranging circles C1 to C3 is minimum. At the initial position shown in FIG. 4, the distance to the two ranging circles C1 and C3 is 0, but the distance to the ranging circle C2 is indicated by arrow L2. Therefore,

[0063] Next, the correction function used in step S2 will be described. The correction function is a function determined using teacher data including the positions of multiple portable devices 103 and the reception strength and ToF-related values ​​at the positions of the multiple portable devices 103, and has a smaller amount of data than the teacher data. The teacher data is, for example, actual measurement data of ToF values ​​and reception strength at 16,530 measurement points inside and outside a vehicle.

[0064] In this embodiment, machine learning capable of handling large amounts of data is used to perform highly accurate correction. Data on the ToF-related values ​​and reception strength of each anchor 30 and portable device 103 is acquired in advance around the vehicle 101 and used for learning. Performing learning that covers the entire vicinity of the vehicle 101 requires handling large amounts of data, but since the logic needs to be executed by the smart ECU 20, which has a small capacity such as an ECU installed in the vehicle 101, a neural network is used as a method for efficient learning, which can reduce the ROM capacity of the smart ECU 20.

[0065] A neural network can reduce the amount of data used compared to pattern matching. As training data, for example, assume that ToF-related values ​​and reception strengths are obtained at 16,530 measurement points inside and outside the vehicle. In this case, ToF-related values ​​and reception strengths are obtained from seven anchors 30 at one measurement point. Therefore, with two parameters, seven antennas, and 16,530 measurement points, 231,420 pieces of data are required. This results in a training data volume of approximately 1,850 KB.

[0066] When performing pattern matching to correct the measurement values ​​acquired by each anchor 30, a large amount of data is preferable, so the training data is used as is and correction is performed by interpolation and extrapolation. In this case, the smart ECU 20 needs to store the training data.

[0067] In contrast, consider the case where a multi-layer perceptron neural network is used, as shown in Figure 5. If the neural network has, for example, an input layer, two hidden layers, and an output layer, the input and output layers will require seven nodes corresponding to the seven anchors 30. 50 nodes are set for each of the two hidden layers. In this case, the thresholds, gains, and weights for each layer will be as shown in Table 1. The number of gains and weights can also be considered the number of paths from each layer to the next layer.

[0068] [Table 1] Therefore, a neural network requires a total of 3,321 pieces of data, including thresholds, gains, and weights. This data volume is approximately 26.5 KB, which is approximately 1 / 70 of the training data. A neural network performs a specified calculation on the input value and outputs a value. Therefore, a neural network can be said to be a correction function.

[0069] Next, another determination method in step S3 of FIG. 3 will be described. When determining whether or not the mobile device 103 is inside the vehicle in step S3, the determination may be made using a statistical model. The statistical model receives the reception strength and ToF-related value acquired in step S1 and outputs whether the mobile device 103 is inside or outside the vehicle. Such a statistical model is trained in advance to determine whether or not the mobile device 103 is inside the vehicle using the same training data as described above. The learning may be performed using machine learning, for example. Deep learning may be used as the machine learning, for example.

[0070] Next, another determination method of step S4 in Fig. 3 will be described. The detailed area determination of step S4 may be performed using another statistical model without using the corrected ToF-related value. Different statistical models may be used for the inside and outside of the vehicle.

[0071] The statistical model for inside the vehicle is configured to input the reception strength and ToF-related values ​​acquired in step S1 and output which of a plurality of pre-set areas inside the vehicle is in which area the portable device 103 is located, thereby identifying the area where the portable device 103 is located. The statistical model for outside the vehicle is configured to input the reception strength and ToF-related values ​​acquired and output which of a plurality of pre-set areas outside the vehicle is in which area the portable device 103 is located, thereby identifying the area where the portable device 103 is located.

[0072] Such a statistical model uses the training data for the vehicle interior to train a statistical model for the vehicle interior, and the training data for the vehicle exterior to train a statistical model for the vehicle exterior. Because the trends in reception strength and ToF-related values ​​differ between the interior and exterior of the vehicle, the area can be identified with greater accuracy.

[0073] As described above, according to the position estimation system and position determination method of this embodiment, the reception strength and ToF-related value of each of the multiple anchors 30 are used to estimate the position of the mobile device 103. The reception strength and the ToF-related value are different parameters, and the factors that cause errors may differ. Therefore, the estimated position of the mobile device 103 when only the reception strength is used may not match the estimated position of the mobile device 103 when only the ToF-related value is used. Therefore, the ToF-related value is corrected using a correction function that inputs the acquired reception strength and ToF-related value and outputs a corrected ToF-related value.

[0074] The correction function is a function determined using teacher data including the positions of the multiple portable devices 103 and the reception strength and ToF-related values ​​at the positions of the multiple portable devices 103. The correction function is generated so that the amount of data is smaller than that of the teacher data. This makes it possible to reduce the amount of data compared to saving all of the teacher data and correcting using the teacher data. Furthermore, since correction is performed using a correction function using two parameters, the correction accuracy can be improved compared to a configuration in which correction is performed using only one of the parameters. As a result, the position of the portable device 103 can be estimated with higher accuracy using the corrected time-of-flight-related values.

[0075] 3, the tentative position of the portable device 103 set using the corrected ToF-related value is set as the initial value of the nonlinear least squares method, and a search is performed from the initial value to estimate the position at which the distance from the corrected time-of-flight-related value is the smallest as the position of the portable device 103. By using the corrected ToF-related value, the initial value can be set with high accuracy, thereby improving the accuracy of position estimation.

[0076] As described above, in this embodiment, not only the ToF-related values ​​of each anchor 30 but also information on radio wave intensity is used as a parameter for estimating the position of the portable device 103. Then, actual measurement data measured in advance at multiple anchors 30 is used to correct the ToF-related values. Machine learning is used for the correction, but a neural network capable of efficient learning is used as a method to ensure that the capacity can be implemented in a microcomputer mounted on the vehicle 101. By applying a neural network, the capacity can be reduced to a level that can be implemented in the smart ECU 20.

[0077] Furthermore, to prevent the key from being locked inside the vehicle and to ensure an area for starting the engine, the system first determines whether the vehicle is inside or outside the vehicle, and then determines the detailed area and estimates the position. By using a neural network, it is possible to determine whether the vehicle is inside or outside with high accuracy. Furthermore, by combining the neural network and selecting the optimal initial position, it is possible to improve the accuracy of positioning using the nonlinear least squares method.

[0078] (Other embodiments) The above describes preferred embodiments of the present disclosure, but the present disclosure is not limited to the above-described embodiments and can be implemented in various modified forms within the scope of the gist of the present disclosure.

[0079] The structures of the above-described embodiments are merely examples, and the scope of the present disclosure is not limited to the scope of these descriptions. The scope of the present disclosure is defined by the claims, and further includes all modifications within the meaning and scope equivalent to the claims.

[0080] In the first embodiment described above, after determining whether the vehicle is inside or outside in step S3 in FIG. 3, the detailed area is determined in step S4. However, the present invention is not limited to this two-stage configuration for determining the detailed area. For example, the area in which the mobile device 103 is located may be identified using a statistical model that inputs the acquired reception strength and ToF-related values ​​and outputs in which of multiple areas set in advance inside and outside the vehicle 101 the mobile device 103 is located. Like other statistical models, the statistical model can be obtained by learning training data. This allows the processes of steps S3 and S4 in FIG. 3 to be combined into one.

[0081] In the first embodiment described above, in step S4 of Fig. 3, one point within the area determined by the detailed area determination is set as the initial value, but the present invention is not limited to this configuration. For example, a ranging circle may be generated from the corrected ToF-related value, and a point equidistant from the intersection of multiple ranging circles may be set as the temporary position, or the center of an area where the intersections of multiple ranging circles are concentrated may be set as the temporary position.

[0082] In the first embodiment described above, the round trip time is used as the ToF-related value, but the present invention is not limited to this configuration. For example, the two-frequency phase difference for each combination of frequencies may be used as the ToF-related value. In BLE communication, since there are two or more frequencies used for communication, two or more two-frequency phase differences for different combinations of frequencies are obtained. The device distance is estimated based on these two-frequency phase differences.

[0083] In a configuration that uses a multi-frequency phase difference as a ToF-related value, communication for ranging can be understood as communication for identifying the transmission and reception phase difference for each of two or more frequencies. Transmitting and receiving CW signals at multiple frequencies can be considered as communication for ranging.

[0084] In the first embodiment described above, the position of the portable device 103 is estimated using the reception strength and the ToF-related value, but other position estimation methods may be combined. For example, the smart ECU 20 may acquire the signal arrival direction as information indicating the reception status of the signal from the portable device 103. The signal arrival direction can be estimated using various methods, such as the MUSIC method or the ESPRIT method. The reception strength, phase, arrival direction, etc. can be called the characteristics of the received signal.

[0085] The functions realized by the smart ECU 20 in the first embodiment may be realized by hardware and software different from those described above, or a combination of these. For example, the smart ECU 20 may communicate with another control device, and the other control device may perform part or all of the processing. When the smart ECU 20 is realized by an electronic circuit, it may be realized by a digital circuit including a large number of logic circuits, or an analog circuit.

[0086] The smart ECU 20 may implement a function as hardware by using one or more ICs. A CPU, MPU, GPU, or DFP (Data Flow Processor) may be used as the processor (computation core). Some or all of the functions of the processor may be implemented by combining multiple types of computational processing devices. Some or all of the functions of the processor may be implemented by using a system-on-chip (SoC), FPGA, ASIC, or the like. FPGA stands for Field-Programmable Gate Array. ASIC stands for Application Specific Integrated Circuit.

[0087] The various computer programs executed by the control device may be stored as computer-executable instructions on a computer-readable non-transitory tangible storage medium, such as a hard-disk drive (HDD), a solid-state drive (SSD), or flash memory.

[0088] In the first embodiment described above, the position estimation device is used in the vehicle 101, but it is not limited to being mounted on the vehicle 101, and at least a part of it may not be mounted on the vehicle 101. [Explanation of symbols]

[0089] 10...Terminal control unit 11...Portable BLE module 12... Portable UWB module 20... Smart ECU (communication control unit, position estimation unit) 21...Processor 22...In-vehicle BLE module 23...In-vehicle UWB module 30...Anchor (communication device) 31... Anchor UWB module 32... Anchor control unit 100...Electronic key system 101...Vehicle 102...In-vehicle system 103...Mobile devices

Claims

1. A position estimation system that estimates a position of a mobile device (103) carried by a user relative to a vehicle (101) by wirelessly communicating with the mobile device, a plurality of communication devices (30) configured to be capable of wirelessly communicating with the portable device and installed at different positions in the vehicle; a communication control unit (20) for controlling the operation of the plurality of communication devices; a location estimation unit (20) that estimates the location of the portable device; The communication control unit acquiring a reception strength of a signal from the mobile device at each of the plurality of communication devices; and acquiring a time-of-flight related value, which is a parameter separate from the reception strength and directly or indirectly indicates a time-of-flight of radio waves from each of the plurality of communication devices to the mobile device; The position estimation unit correcting the time-of-flight related value using a correction function that receives the acquired reception intensity and the time-of-flight related value as input and outputs the corrected time-of-flight related value; estimating a position of the mobile device using the corrected time-of-flight related value; The correction function is a function determined using training data including the positions of multiple mobile devices and the reception strength and flight time-related values ​​at the positions of multiple mobile devices, and the position estimation system has a smaller amount of data than the training data.

2. The position estimation unit setting the tentative position of the mobile device, which is set using the corrected time-of-flight related value, as an initial value for a nonlinear least squares method; The position estimation system according to claim 1 , wherein a search is performed from the initial value, and a position where a distance determined from the corrected time-of-flight related value is the smallest is estimated as the position of the mobile device.

3. The position estimation unit Identifying the area where the mobile device is located using a statistical model that outputs, by inputting the acquired reception strength and the flight time related value, in which of a plurality of areas that are preset inside and outside the vehicle the mobile device is located; A point within the specified area is set as the initial value for the nonlinear least squares method, The position estimation system according to claim 1 , wherein a search is performed from the initial value, and a position where a distance determined from the corrected time-of-flight related value is the smallest is estimated as the position of the mobile device.

4. The position estimation unit determining whether the portable device is located inside the vehicle using at least one of the acquired reception strength and the acquired time-of-flight related value; If the portable device is located inside the vehicle, the area where the portable device is located is identified using a statistical model that outputs an area where the portable device is located among a plurality of pre-defined areas inside the vehicle by inputting the acquired reception strength and the flight time-related value; If the portable device is outside the vehicle cabin, the area where the portable device is located is identified using a statistical model that outputs an area where the portable device is located among a plurality of pre-defined areas outside the vehicle cabin by inputting the acquired reception strength and the flight time-related value; A point within the specified area is set as the initial value for the nonlinear least squares method, The position estimation system according to claim 1 , wherein a search is performed from the initial value, and a position where a distance determined from the corrected time-of-flight related value is the smallest is estimated as the position of the mobile device.

5. 1. A location estimation method executed by at least one processor for estimating a location of a mobile device carried by a user, the method comprising: acquiring a reception strength of a signal from the portable device from a plurality of communication devices that are configured to be able to wirelessly communicate with the portable device and that are disposed at different positions in the vehicle; acquiring a time-of-flight related value, which is a parameter separate from the reception strength and directly or indirectly indicates a time-of-flight of radio waves from each of the plurality of communication devices to the mobile device; correcting the time-of-flight related value using a correction function that receives the acquired reception intensity and the time-of-flight related value as input and outputs the corrected time-of-flight related value; and estimating a position of the mobile device using the corrected time-of-flight related value; The correction function is a function determined using training data including the positions of multiple mobile devices and the reception strength and flight time-related values ​​at the positions of multiple mobile devices, and is a position estimation method in which the amount of data is smaller than that of training data.

Images