Vehicle load measuring method and system based on multi-sensor data fusion technology

Abstract

The invention provides a vehicle load measurement method and system based on a multi-sensor data fusion technology, and belongs to the field of measurement. According to the invention, various sensors are arranged on the vehicle, the vehicle load is obtained in real time based on the multi-sensor data fusion technology, the load data is uploaded to the management platform and the user side through the vehicle-mounted communication module, the vehicle overload condition can be dynamically obtained, and many problems existing in a static measurement method are solved. Meanwhile, when the vehicle load is estimated, external quality factors and internal engine load factors are comprehensively considered, so that load measurement is more accurate, and overload early warning is more timely and effective. Furthermore, a machine learning method is introduced to improve the accuracy of data output, and a machine learning algorithm is optimized based on complex conditions during dynamic driving of the vehicle, so that the applicability of the algorithm is improved.

Description

Technical Field

[0001] The invention belongs to the field of measurement, and in particular relates to a vehicle load measurement method and system based on multi-sensor data fusion technology. Background Art

[0002] Usually, in order to improve transportation efficiency and economic benefits, vehicle overloading is a common phenomenon everywhere. However, overloaded transportation not only causes huge damage to the public transportation infrastructure, but also poses a serious threat to people’s property and life safety. How to solve the problem of vehicle overloading and over-limit is imminent.

[0003] At present, my country mainly controls vehicle overloading by setting up monitoring stations. This control method requires setting up monitoring stations along the highway, guiding vehicles on the road to pass through the weighing equipment at the monitoring stations, and using static weighing to determine whether the vehicles are overloaded. However, the monitoring stations are fixed in position, which makes it easy for vehicles to escape monitoring, and it is easy to cause traffic congestion when the traffic volume is large, making the monitoring efficiency low.

[0004] In order to solve the problems caused by static weighing, a vehicle-mounted dynamic weighing technology has been proposed. It uses displacement sensors to measure the tiny deformation of the frame or axle and converts it into vehicle load, thereby realizing real-time weighing of the vehicle during driving, improving monitoring efficiency and avoiding traffic congestion.

[0005] However, the vehicle condition and road conditions are changeable during driving. The age of the vehicle, road bumps, and weather conditions will affect the accuracy of the sensor. This makes the current vehicle dynamic weighing technology unable to accurately obtain the vehicle load weight, and the monitoring credibility is not high. Summary of the invention

[0006] The purpose of the present invention is to solve the above-mentioned problems existing in the prior art and to provide a method and system capable of accurately realizing dynamic measurement of vehicle load.

[0007] The present invention provides a vehicle load measurement method based on multi-sensor data fusion technology, comprising the following steps:

[0008] 1) Data collection: sensors are placed on the vehicle to collect vehicle data, including but not limited to driving speed, acceleration, driving force, and driving tilt angle;

[0009] 2) Data transmission: The collected data is transmitted to the vehicle control module in real time;

[0010] 3) Data processing: The vehicle control module processes the data and calculates the vehicle load, including the following steps:

[0011] (31) Using formula (1), the external mass of the vehicle is estimated:

[0012]

[0013] Among them, m represents the external mass of the vehicle, the unit is kg, F represents the driving force of the vehicle, the unit is N, R represents the resistance of the vehicle, the unit is N, and a represents the acceleration of the vehicle, the unit is m / s 2 , g represents the acceleration due to gravity, the unit is m / s 2 , Θ represents the inclination angle of the vehicle;

[0014] (32) Formula (2) is used to estimate the vehicle internal load caused by the engine factor,

[0015]

[0016] Wherein, E represents the internal load of the vehicle, C represents the current engine intake air flow, P represents the peak air flow when the throttle is fully opened under standard temperature and pressure, B represents the mercury column pressure, and A represents the air temperature;

[0017] (33) The real-time vehicle load is affected by internal factors (formula (2)) and external factors (formula (1)). Formula (3) is used to calculate the real-time vehicle load.

[0018]

[0019] Where W represents the vehicle load, b represents the offset value, and x i represents the i-th characteristic variable, k i Represents the characteristic coefficient of the i-th characteristic variable, with a total of n characteristics taken into account;

[0020] 4) Model optimization: Use machine learning algorithms to optimize the data processing process to obtain more accurate vehicle load conditions;

[0021] 5) Data upload: The vehicle control module uploads the final vehicle load condition to the management platform through the communication module;

[0022] 6) Data display: The management platform sends the data to the vehicle user terminal.

[0023] Furthermore, the following models are used to optimize the machine learning algorithm, including:

[0024]

[0025] Among them, y represents the dependent variable in the model, β0 represents the intercept term, and β n Represents the nth feature variable x n The characteristic coefficient of , a total of N features are taken into account, ε represents random noise;

[0026] In order to obtain the best characteristic coefficient β, the function in formula (5)-(6) is used for estimation:

[0027]

[0028] Among them, RSS represents the sum of squared errors, y n represents a known actual value, Represents the output value of the model; Represents the characteristic coefficient of the final output of the model, Y=(y1,...,y N ) T represents a vector matrix, X is the covariate matrix, X T is the transpose of the covariate matrix.

[0029] Furthermore, the sensor includes but is not limited to an acceleration sensor, a gyroscope, a temperature sensor, a pressure sensor, and a displacement sensor.

[0030] Furthermore, the vehicle-mounted user terminal is also provided with an early warning module, which sends an overload early warning to the user when the vehicle load weight exceeds a threshold.

[0031] The present invention also provides a vehicle load measurement system based on multi-sensor data fusion technology, comprising:

[0032] 1) Data acquisition module: multiple sensors are arranged on the vehicle, and the sensors collect vehicle data, including but not limited to driving speed, acceleration, driving force, and driving tilt angle;

[0033] 2) Data transmission module: transmits the collected data to the vehicle control module in real time;

[0034] 3) Vehicle control module: The vehicle control module processes the data and calculates the vehicle load, including the following steps:

[0035] (31) Using formula (1), the external mass of the vehicle is estimated:

[0036]

[0037] Among them, m represents the external mass of the vehicle, the unit is kg, F represents the driving force of the vehicle, the unit is N, R represents the resistance of the vehicle, the unit is N, and a represents the acceleration of the vehicle, the unit is m / s 2 , g represents the acceleration due to gravity, the unit is m / s 2 , Θ represents the inclination angle of the vehicle;

[0038] (32) Formula (2) is used to estimate the vehicle internal load caused by the engine factor,

[0039]

[0040] Wherein, E represents the internal load of the vehicle, C represents the current engine intake air flow, P represents the peak air flow when the throttle is fully opened under standard temperature and pressure, B represents the mercury column pressure, and A represents the air temperature;

[0041] (33) The real-time vehicle load is affected by internal factors (formula (2)) and external factors (formula (1)). Formula (3) is used to calculate the real-time vehicle load.

[0042]

[0043] Where W represents the vehicle load, b represents the offset value, and x i represents the i-th characteristic variable, k i Represents the characteristic coefficient of the i-th characteristic variable, with a total of n characteristics taken into account;

[0044] 4) Communication module: The vehicle control module uploads the final vehicle load condition to the management platform through the communication module;

[0045] 5) In-vehicle user terminal: The management platform sends data to the in-vehicle user terminal.

[0046] Furthermore, the vehicle control module also includes a model optimization module, and the model optimization module optimizes the data processing process using a machine learning algorithm, and the algorithm includes:

[0047]

[0048] Among them, y represents the dependent variable in the model, β0 represents the intercept term, and β n Represents the nth feature variable x n The characteristic coefficient of , a total of N features are taken into account, ε represents random noise;

[0049] In order to obtain the best characteristic coefficient β, the function in formula (5)-(6) is used for estimation:

[0050]

[0051] Among them, RSS represents the sum of squared errors, y n represents a known actual value, Represents the output value of the model; Represents the characteristic coefficient of the final output of the model, Y=(y1,...,y N ) T represents a vector matrix, X is the covariate matrix, X T is the transpose of the covariate matrix.

[0052] Furthermore, the sensor includes but is not limited to an acceleration sensor, a gyroscope, a temperature sensor, a pressure sensor, and a displacement sensor.

[0053] Furthermore, the vehicle-mounted user terminal is also provided with an early warning module, which sends an overload early warning to the user when the vehicle load weight exceeds a threshold.

[0054] Compared with the prior art, the beneficial effects of the present invention are as follows: by deploying a variety of sensors on the vehicle, the present invention acquires the vehicle load in real time based on multi-sensor data fusion technology, and uploads the load data to the management platform and the user end through the vehicle-mounted communication module, so as to dynamically acquire the vehicle overload situation and solve many problems existing in the static measurement method. At the same time, when estimating the vehicle load, the external quality factors and the internal engine load factors are comprehensively considered, so that the load measurement is more accurate and the overload warning is more timely and effective. Furthermore, the machine learning method is introduced to improve the accuracy of data output, and the machine learning algorithm is optimized based on the complex situation when the vehicle is driving dynamically to improve the applicability of the algorithm. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 It is a flow chart of a vehicle load measurement method based on multi-sensor data fusion technology provided by the present invention.

[0056] Figure 2 It is a schematic diagram of a vehicle load measurement system based on multi-sensor data fusion technology provided by the present invention. DETAILED DESCRIPTION

[0057] The present invention is further described in detail below in conjunction with the accompanying drawings:

[0058] Attached Figure 1 The present invention provides a vehicle load measurement method based on multi-sensor data fusion technology, comprising the following steps:

[0059] 1) Data collection: sensors are placed on the vehicle to collect vehicle data, including but not limited to driving speed, acceleration, driving force, and driving tilt angle;

[0060] 2) Data transmission: The collected data is transmitted to the vehicle control module in real time;

[0061] 3) Data processing: the vehicle control terminal processes the data and calculates the vehicle load;

[0062] 4) Model optimization: Use machine learning algorithms to optimize the data processing process to obtain more accurate vehicle load conditions;

[0063] 5) Data upload: The vehicle control module uploads the final vehicle load condition to the management platform through the communication module;

[0064] 6) Data display: The management platform sends the data to the vehicle user terminal.

[0065] Among them, the method of calculating vehicle load includes:

[0066] 31) Use formula (1) to estimate the external mass of the vehicle,

[0067]

[0068] Among them, m represents the external mass of the vehicle, the unit is kg, F represents the driving force of the vehicle, the unit is N, R represents the resistance of the vehicle, the unit is N, and a represents the acceleration of the vehicle, the unit is m / s 2 , g represents the acceleration due to gravity, the unit is m / s 2 , Θ represents the inclination angle of the vehicle;

[0069] 32) Use formula (2) to estimate the vehicle internal load caused by engine factors,

[0070]

[0071] Wherein, E represents the internal load of the vehicle, C represents the current engine intake air flow, P represents the peak air flow when the throttle is fully opened under standard temperature and pressure, B represents the mercury column pressure, and A represents the air temperature;

[0072] 33) The real-time vehicle load is affected by internal factors (Formula (2)) and external factors (Formula (1)). We believe that this relationship is a multivariate linear regression, so Formula (3) is used to calculate the real-time vehicle load.

[0073]

[0074] Where W represents the vehicle load, b represents the offset value, and x i represents the i-th characteristic variable, k i Represents the characteristic coefficient of the i-th characteristic variable, with a total of n characteristics taken into account.

[0075] Furthermore, the following models are used to optimize the machine learning algorithm, including:

[0076]

[0077] Among them, y represents the dependent variable in the model, β0 represents the intercept term, and β n Represents the nth feature variable x n The characteristic coefficient of , a total of N features are taken into account, ε represents random noise;

[0078] In order to obtain the best characteristic coefficient β, the function in formula (5)-(6) is used for estimation:

[0079]

[0080] Among them, RSS represents the sum of the square error, y n represents a known actual value, Represents the output value of the model; Represents the characteristic coefficient of the final output of the model, Y=(y1,...,y N ) T represents a vector matrix, X is the covariate matrix, X T is the transpose of the covariate matrix.

[0081] In order to obtain vehicle data more accurately, the deployed sensors include but are not limited to acceleration sensors, gyroscopes, temperature sensors, pressure sensors, and displacement sensors.

[0082] The vehicle-mounted user terminal is also provided with an early warning module, which sends an overload early warning to the user when the vehicle load weight exceeds a threshold.

[0083] The present invention also provides a vehicle load measurement system based on multi-sensor data fusion technology, the schematic diagram of the system is shown in Figure 2 As shown, specifically including:

[0084] 1) Data acquisition module: multiple sensors are arranged on the vehicle, and the sensors collect vehicle data, including but not limited to driving speed, acceleration, driving force, and driving tilt angle;

[0085] 2) Data transmission module: transmits the collected data to the vehicle control module in real time;

[0086] 3) Vehicle control module: The vehicle control module processes the data and calculates the vehicle load, including the following steps:

[0087] (31) Using formula (1), the external mass of the vehicle is estimated:

[0088]

[0089] Among them, m represents the external mass of the vehicle, the unit is kg, F represents the driving force of the vehicle, the unit is N, R represents the resistance of the vehicle, the unit is N, and a represents the acceleration of the vehicle, the unit is m / s 2 , g represents the acceleration due to gravity, the unit is m / s 2 , Θ represents the inclination angle of the vehicle;

[0090] (32) Formula (2) is used to estimate the vehicle internal load caused by the engine factor,

[0091]

[0092] Wherein, E represents the internal load of the vehicle, C represents the current engine intake air flow, P represents the peak air flow when the throttle is fully opened under standard temperature and pressure, B represents the mercury column pressure, and A represents the air temperature;

[0093] (33) The real-time vehicle load is affected by internal factors (formula (2)) and external factors (formula (1)). Formula (3) is used to calculate the real-time vehicle load.

[0094]

[0095] Where W represents the vehicle load, b represents the offset value, and x i represents the i-th characteristic variable, k i Represents the characteristic coefficient of the i-th characteristic variable, with a total of n characteristics taken into account;

[0096] 4) Communication module: The vehicle control module uploads the final vehicle load condition to the management platform through the communication module;

[0097] 5) In-vehicle user terminal: The management platform sends data to the in-vehicle user terminal.

[0098] Furthermore, the vehicle control module also includes a model optimization module, and the model optimization module optimizes the data processing process using a machine learning algorithm, and the algorithm includes:

[0099]

[0100] Among them, y represents the dependent variable in the model, β0 represents the intercept term, and β n Represents the nth feature variable x n The characteristic coefficient of , a total of N features are taken into account, ε represents random noise;

[0101] In order to obtain the best characteristic coefficient β, the function in formula (5)-(6) is used for estimation:

[0102]

[0103] Among them, RSS represents the sum of squared errors, y n represents a known actual value, Represents the output value of the model; Represents the characteristic coefficient of the final output of the model, Y=(y1,...,y N ) Trepresents a vector matrix, X is the covariate matrix, X T is the transpose of the covariate matrix.

[0104] Furthermore, the sensor includes but is not limited to an acceleration sensor, a gyroscope, a temperature sensor, a pressure sensor, and a displacement sensor.

[0105] Furthermore, the vehicle-mounted user terminal is also provided with an early warning module, which sends an overload early warning to the user when the vehicle load weight exceeds a threshold.

[0106] The present invention can dynamically obtain the vehicle overload situation by deploying multiple sensors on the vehicle, acquiring the vehicle load in real time based on multi-sensor data fusion technology, and uploading the load data to the management platform and the user end through the vehicle-mounted communication module, thereby solving many problems existing in the static measurement method. At the same time, when estimating the vehicle load, the external quality factors and the internal engine load factors are comprehensively considered, so that the load measurement is more accurate and the overload warning is more timely and effective. Furthermore, the machine learning method is introduced to improve the accuracy of data output, and the machine learning algorithm is optimized based on the complex situation when the vehicle is driving dynamically to improve the applicability of the algorithm.

[0107] In the description of the present invention, it should be noted that, unless otherwise clearly specified and limited, the terms "connected" and "connection" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0108] In the description of the present invention, unless otherwise specified, the terms "upper", "lower", "left", "right", "inside", "outside", etc. indicate directions or positional relationships based on the directions or positional relationships shown in the accompanying drawings. They are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific direction, be constructed and operated in a specific direction. Therefore, they cannot be understood as limitations on the present invention.

[0109] Finally, it should be explained that the above technical solution is only one implementation method of the present invention. For those skilled in the art, it is easy to make various types of improvements or modifications based on the application methods and principles disclosed in the present invention, and it is not limited to the method described in the above specific implementation method of the present invention. Therefore, the method described above is only preferred and does not have a restrictive meaning.

Claims

1. A vehicle load measurement method based on multi-sensor data fusion technology, comprising the following steps: 1) Data collection: sensors are placed on the vehicle to collect vehicle data, including but not limited to driving speed, acceleration, driving force, and driving tilt angle; 2) Data transmission: The collected data is transmitted to the vehicle control module in real time; 3) Data processing: The vehicle control module processes the data and calculates the vehicle load, including the following steps: (31) Using formula (1), the external mass of the vehicle is estimated: Among them, m represents the external mass of the vehicle, the unit is kg, F represents the driving force of the vehicle, the unit is N, R represents the resistance of the vehicle, the unit is N, and a represents the acceleration of the vehicle, the unit is m / s 2 , g represents the acceleration due to gravity, the unit is m / s 2 , Θ represents the inclination angle of the vehicle; (32) Formula (2) is used to estimate the vehicle internal load caused by the engine factor, Wherein, E represents the internal load of the vehicle, C represents the current engine intake air flow, P represents the peak air flow when the throttle is fully opened under standard temperature and pressure, B represents the mercury column pressure, and A represents the air temperature; (33) The real-time vehicle load is affected by internal factors (formula (2)) and external factors (formula (1)). Formula (3) is used to calculate the real-time vehicle load. Where W represents the vehicle load, b represents the offset value, and x i represents the i-th feature variable, k i Represents the characteristic coefficient of the i-th characteristic variable, with a total of n characteristics taken into account; 4) Model optimization: Use machine learning algorithms to optimize the data processing process to obtain more accurate vehicle load conditions; 5) Data upload: The vehicle control module uploads the final vehicle load condition to the management platform through the communication module; 6) Data display: The management platform sends the data to the vehicle user terminal.

2. The vehicle load measurement method based on multi-sensor data fusion technology according to claim 1 is characterized in that: The following models are used to optimize machine learning algorithms, including: Among them, y represents the dependent variable in the model, β0 represents the intercept term, and β n Represents the nth feature variable x n The characteristic coefficient of , a total of N features are taken into account, ε represents random noise; In order to obtain the best characteristic coefficient β, the function in formula (5)-(6) is used for estimation: Among them, RSS represents the sum of squared errors, y n represents a known actual value, Represents the output value of the model; Represents the characteristic coefficient of the final output of the model, Y=(y1,...,y N ) T represents a vector matrix, X is the covariate matrix, X T is the transpose of the covariate matrix.

3. The vehicle load measurement method based on multi-sensor data fusion technology according to claim 1 is characterized in that: The sensor includes but is not limited to an acceleration sensor, a gyroscope, a temperature sensor, a pressure sensor, and a displacement sensor.

4. The vehicle load measurement method based on multi-sensor data fusion technology according to claim 1 is characterized in that: The vehicle-mounted user terminal is also provided with an early warning module, which sends an overload early warning to the user when the vehicle load weight exceeds a threshold.

5. A vehicle load measurement system based on multi-sensor data fusion technology, comprising: 1) Data acquisition module: multiple sensors are arranged on the vehicle, and the sensors collect vehicle data, including but not limited to driving speed, acceleration, driving force, and driving tilt angle; 2) Data transmission module: transmits the collected data to the vehicle control module in real time; 3) On-board control module: The on-board control module processes the data and calculates the vehicle load, including the following steps: (31) Using formula (1), the external mass of the vehicle is estimated: Among them, m represents the external mass of the vehicle, the unit is kg, F represents the driving force of the vehicle, the unit is N, R represents the resistance of the vehicle, the unit is N, and a represents the acceleration of the vehicle, the unit is m / s 2 , g represents the acceleration due to gravity, the unit is m / s 2 , Θ represents the inclination angle of the vehicle; (32) Formula (2) is used to estimate the vehicle internal load caused by the engine factor, Wherein, E represents the internal load of the vehicle, C represents the current engine intake air flow, P represents the peak air flow when the throttle is fully opened under standard temperature and pressure, B represents the mercury column pressure, and A represents the air temperature; (33) The real-time vehicle load is affected by internal factors (formula (2)) and external factors (formula (1)). Formula (3) is used to calculate the real-time vehicle load. Where W represents the vehicle load, b represents the offset value, and x i represents the i-th feature variable, k i Represents the characteristic coefficient of the i-th characteristic variable, with a total of n characteristics taken into account; 4) Communication module: The vehicle control module uploads the final vehicle load condition to the management platform through the communication module; 5) In-vehicle user terminal: The management platform sends data to the in-vehicle user terminal.

6. The vehicle load measurement system based on multi-sensor data fusion technology according to claim 5 is characterized in that: The vehicle control module also includes a model optimization module, which uses a machine learning algorithm to optimize the data processing process. The algorithm includes: Among them, y represents the dependent variable in the model, β0 represents the intercept term, and β n Represents the nth feature variable x n The characteristic coefficient of , a total of N features are taken into account, ε represents random noise; In order to obtain the best characteristic coefficient β, the function in formula (5)-(6) is used for estimation: Among them, RSS represents the sum of squared errors, y n represents a known actual value, Represents the output value of the model; Represents the characteristic coefficient of the final output of the model, Y=(y1,...,y N ) T represents a vector matrix, X is the covariate matrix, X T is the transpose of the covariate matrix.

7. The vehicle load measurement system based on multi-sensor data fusion technology according to claim 5 is characterized in that: The sensor includes but is not limited to an acceleration sensor, a gyroscope, a temperature sensor, a pressure sensor, and a displacement sensor.

8. The vehicle load measurement system based on multi-sensor data fusion technology according to claim 5, characterized in that: The vehicle-mounted user terminal is also provided with an early warning module, which sends an overload early warning to the user when the vehicle load weight exceeds a threshold.

Images