Method for determining two-person use of an electric scooter

Abstract

The proposed invention relates to methods for determining two-person use of an electric scooter. The present method includes obtaining data from sensors mounted on an electric scooter, and more specifically: wheel speed, motor current, scooter charge, pressure on the throttle, a right brake lever value, and a left brake lever value. During a predetermined period of time, the following values are determined: the total time for which sensor data is available; the mean time for which sensor data is available; the standard deviation of the time for which data is available; the minimum and maximum time for which data were available, and the 25th, 50th and 75th percentiles of the data distribution. The obtained data and values are input into a machine learning model, at the output of which a probability of two-person use is obtained, where 0 is single-person use and 1 is two-person use. The technical result consists in determining two-person use of an electric scooter.

Description

[0001] A way to determine whether two people will travel on an electric scooter

[0002] AREA OF TECHNOLOGY

[0003] This technical solution relates to vehicles, namely to methods for detecting two-person trips on an electric scooter.

[0004] LEVEL OF TECHNOLOGY

[0005] Efficient and safe use of an electric scooter is an important rule for all users. However, riding with two people on an electric scooter presents a number of serious challenges and risks. Electric scooters are typically designed for single occupants. A passenger, clinging to or holding onto the driver, can disrupt the balance and stability of the scooter, potentially leading to loss of control and an accident. Controlling an electric scooter with a passenger becomes significantly more difficult due to the shift in the scooter's center of gravity, making maneuvering difficult, especially when turning or avoiding obstacles. Also, doubling the load on the scooter can lead to accelerated wear and tear, which can reduce its lifespan and require additional repair or replacement costs.

[0006] Currently, monitoring double rides is only possible visually, but this monitoring is ineffective. The proposed solution develops a method for detecting double rides on an electric scooter by processing data received from the scooter's sensor.

[0007] The prior art includes solutions aimed at limiting the speed of personal mobility vehicles when in a certain zone, which process data from sensors (US11034404 (B2), published 2021-06-15, US10757485 (B2), published 2020-08-25).

[0008] However, these solutions do not solve the problem of identifying two-person trips on an electric scooter.

[0009] ESSENCE OF THE INVENTION

[0010] The technical challenge addressed by the proposed solution is to create a method for detecting two-person trips on an electric scooter by processing data received from the scooter's sensors.

[0011] The technical result achieved by solving the above technical problem is the definition of trips for two on an electric scooter and coincides with the stated technical problem.The claimed technical result is achieved by implementing a method for determining trips by two people on an electric scooter, which includes the following stages: receiving data from sensors installed on the electric scooter, namely: wheel rotation speed, motor current, electric scooter charge, gas trigger pressing value, right brake handle value, left brake handle value; over a predetermined period of time, the following values ​​are determined: total time of presence of data received from the sensors; average time of presence of data received from the sensors; standard deviation of the time of presence of data; minimum and maximum time for which the data was present, data distribution percentiles - 25%, 50%, 75%; the obtained data and values ​​are fed to the input of a machine learning model, the output of which is the probability of a trip being made by two people, where 0 means the trip was made by one user, 1 means the trip was made by two people.

[0012] DETAILED DESCRIPTION OF THE INVENTION

[0013] The following detailed description of the invention includes numerous implementation details to provide a clear understanding of the present invention. However, one skilled in the art will readily understand how the present invention may be used with or without these implementation details. In other instances, well-known methods, procedures, and components have not been described in detail to avoid obscuring the features of the present invention.

[0014] Furthermore, it will be clear from the foregoing description that the invention is not limited to the embodiment described. Numerous possible modifications, changes, variations, and substitutions, while preserving the spirit and form of the present invention, will be apparent to those skilled in the art.

[0015] The proposed solution identifies the difference in a person's behavior when he is riding alone and when a passenger is riding with him on the same platform.

[0016] A scooter is a complex and unstable device with two points of contact with the ground. A solo passenger can balance freely and effortlessly on the scooter, taking into account the specific conditions of the road, surrounding terrain, and speed. However, with a second person on the platform, controlling the scooter becomes more difficult due to factors such as limited freedom of movement, instability due to the second person also attempting to balance on the platform, limited visibility, and other factors. Consequently, people riding together on the same scooter tend to be unstable and overly careful (compared to people riding alone).

[0017] For example, tandem riders rarely press the throttle fully, and press the brake levers more often. Battery consumption is increased, and the current to the motor wheel is at its peak.

[0018] Moreover, the model takes these factors into account holistically. That is, the model doesn't determine that a ride was shared simply because the user braked more frequently. It takes all 72 features together and compares, in this example, brake pressure with the battery drain and the current flowing through the hub motor.

[0019] The method for determining whether two people are traveling on an electric scooter is as follows.

[0020] The moment the electric scooter is rented, it begins recording data from the installed sensors into its memory.

[0021] The sensors used to collect information and which are installed on the electric scooter are: a barometer, an accelerometer, a gyroscope, a motor-wheel speed sensor, a battery current and voltage sensor.

[0022] Based on the data collected from the sensors, the following parameters are sent to the electric scooter's UT module with a resolution of 100 ms when the wheel rotation speed is not equal to 0:

[0023] • Operating time of the electric scooter from the moment the electric scooter is loaded, in s

[0024] • wheel speed in m / s

[0025] • maximum wheel rotation speed for the zone in m / s

[0026] • battery current, A

[0027] • electric scooter charge, in %

[0028] • atmospheric pressure according to the barometer, hPa

[0029] • normalized throttle trigger (0 - 100)

[0030] • right brake handle (0 - 100)

[0031] • left brake handle (0 - 100)

[0032] • Euler angles, degrees

[0033] The received data is converted into values ​​for data processing.

[0034] Each received value is processed using each of the following algorithms:

[0035] 1. Average value, code suffix "mean" 2. Standard deviation hU s: / / en.wjk(p^ code suffix "std"

[0036] 3. Minimum value, code suffix "min"

[0037] 4. 25th percentile code suffix "25"

[0038] 5. 50th percentile (median), code suffix "50"

[0039] 6. 75th percentile, code suffix "75"

[0040] 7. Maximum value, code suffix "max"

[0041] 8. Sums, code suffix "sum"

[0042] Calculations are performed based on the accumulated 20 units of data - approximately for the past 2 seconds (data from the sensors is received with a resolution of 100 Ohms, when the speed of the motor-wheel (wh) is not equal to 0, accordingly, the above algorithms are calculated based on 20 units of data from the sensors; 20 units of data from the sensors is not equal to data for 2 seconds, since in the middle of the 2-second interval there is a possibility that the electric vehicle is standing still).

[0043] At the output, 72 features are obtained (nine parameters (wheel rotation speed; maximum wheel rotation speed for the zone; battery current; electric scooter charge; atmospheric pressure according to the barometer; normalized throttle trigger; right brake handle; left brake handle; Euler angles, processed using the eight algorithms mentioned above). Each feature is designated as <name of sensor data>_<code suffix.

[0044] This data is sent from the electric scooter to the server, where it is processed to determine whether two-person trips are taking place.

[0045] Example of description of features

[0046] Let's take as an example wh - speed in m / s and ch - charge in % using 11 data units (see Table 1):

[0047] Table 1

[0048] Only data from 1 to 3 and from 5 to 11 will be taken into account, since at this moment the speed is not uniform. Then, for example, the average values ​​will be calculated as:

[0049] • for speed: Then the "wh_mean" feature for this data section will be equal to 5, and ch mean = 52.3

[0050] Machine learning models are trained using the gradient boosting algorithm.

[0051] To train the algorithm, the initial labeled trip data is divided into train and test trips in a ratio of 70 to 30.

[0052] Train is the data used to train the model, i.e., to find the weights and coefficients that most accurately describe the feature space with a single formula. Test is the control data used to evaluate whether the algorithm was able to find a generalizing function (i.e., whether the algorithm was able to predict whether a trip was two-person or not).

[0053] How gradient boosting works for classification.

[0054] Detecting two-person trips is a classification problem, and the algorithm model itself answers the question "How likely is it that during the specified time interval (~2 s) the electric scooter was carrying two passengers?"

[0055] X is a table containing all the features (72 features) of an electric scooter based on trips. y is a column vector that describes the feature of a single (y = 0) or double (y = 1) trip.

[0056] X train, X test is the table X, divided into train and test data, respectively. y_train, y test is the column vector y, divided into train and test data, respectively.

[0057] The algorithm works as follows.

[0058] 1. For y_train and the initial forecast, values ​​0 are assigned for each of the 72 features;

[0059] 2. Forecasts are transformed into probabilities using the softmax function;

[0060] 3. The residuals of the model are calculated based on the antigradient of the loss and probability functions;

[0061] 4. The regression tree is trained on X_train and residuals, then a forecast is made on X_train;

[0062] 5. For each leaf in the tree, coefficients are calculated based on the residuals taken from the positions of observations that fall within a specific leaf node;

[0063] 6. The obtained forecasts for each class and the sum of the coefficients are added to the original ones;

[0064] 7. Steps 2-6 are repeated for each tree in each feature;

[0065] 8. After all models have been trained, the initial forecast from step 1 is created;

[0066] 9. Next, predictions are made for X_test on the trained trees for each class and added to the original ones;

[0067] 10. The classes with the highest sum will be the final forecast.

[0068] While the scooter is moving, data is collected and processed as described above. If the machine learning model outputs a value of 1 (meaning the trip was completed by two people), the scooter emits a sound signal indicating that the speed limit is about to be applied and a command to slow the scooter to a speed no higher than 10 km / h.

[0069] A computing system that provides the data processing necessary for the implementation of the claimed solution generally contains the following components: one or more processors, at least one memory, a data storage means, input / output interfaces, an input means, and network interaction means.

[0070] When executing machine-readable commands contained in the RAM, the processor of the device is configured to perform the basic computing operations necessary for the operation of the device or the functionality of one or more of its components.

[0071] Memory is typically implemented as RAM, where the necessary software logic is loaded to provide the required functionality. When implementing the proposed solution, the memory capacity required for its implementation is allocated.

[0072] The data storage device can be a HDD, SSD, RAID array, network storage, flash memory, etc. It enables long-term storage of various types of information, such as the aforementioned files with user / passenger data sets, databases containing records of time intervals measured for each user, user IDs, etc.

[0073] Interfaces are standard means for connecting and operating peripherals and other devices, such as USB, RS232, RJ45, COM, HDMI, PS / 2, Lightning, etc.

[0074] The choice of interfaces depends on the specific device design, which may be a personal computer, mainframe, server cluster, thin client, smartphone, laptop, etc.

[0075] A keyboard may be used as a data input device in any embodiment of the system implementing the described method. The keyboard hardware may be any known device: it could be a built-in keyboard used on a laptop or netbook, or a separate device connected to a desktop computer, server, or other computing device. The connection may be either wired, in which the keyboard cable is connected to a PS / 2 or USB port located on the desktop computer's system unit, or wireless, in which the keyboard exchanges data via a wireless channel, such as a radio channel, with a base station, which in turn is directly connected to the system unit, for example, to one of the USB ports.In addition to the keyboard, data input devices may also include: a joystick, display (touch screen), projector, touchpad, mouse, trackball, light pen, speakers, microphone, etc.

[0076] Network communication tools are selected from a device that provides network data reception and transmission, such as an Ethernet card, WLAN / Wi-Fi module, Bluetooth module, BLE module, NFC module, IrDA, RFID module, GSM modem, etc. These tools enable data exchange via a wired or wireless data transmission channel, such as WAN, PAN, LAN, Intranet, Internet, WLAN, WMAN, or GSM.

[0077] The device components are connected via a common data bus.

[0078] In these application materials, a preferred disclosure of the implementation of the claimed technical solution was presented, which should not be used as limiting other particular embodiments of its implementation, which do not go beyond the scope of the requested scope of legal protection and are obvious to specialists in the relevant field of technology.

Claims

FORMULA 1. A method for detecting trips by two people on an electric scooter, which comprises the following steps: receiving data from sensors installed on the electric scooter, namely: wheel rotation speed, motor current, electric scooter charge, throttle trigger press value, right brake handle value, left brake handle value; over a predetermined period of time, determining the following values: total time of presence of data received from the sensors; average time of presence of data received from the sensors; standard deviation of the time of presence of data; minimum and maximum time for which the data was present, data distribution percentiles - 25%, 50%, 75%; the obtained data and values ​​are fed to the input of a machine learning model, the output of which is the probability of a trip being completed by two people, where 0 means the trip was completed by one user, 1 means the trip was completed by two people.

Images