Electric bicycle state monitoring method based on sensor fusion

Abstract

The invention discloses an electric bicycle state monitoring method based on sensor fusion, and relates to the technical field of data analysis, and the method comprises the steps: obtaining a resonance cooperation index and a road energy permeability index through multi-dimensional data calculation, and carrying out the fusion to obtain a global feature vector; inputting the global feature vector under the time sequence into a preset double-branch prediction model, and outputting a riding intention and a riding mode; calculating a power demand according to the multi-dimensional data, performing adaptive power distribution according to the power demand and the riding mode, and when the riding intention is braking or downhill, performing energy recovery according to a preset energy recovery strategy; and calculating predicted energy consumption and a cost index of the road section, and performing path planning. According to the method, cooperation of a rider and a motor and road surface conditions are quantified through a resonance cooperation index and a road energy permeability index, riding intention and mode prediction is achieved through a double-branch prediction model, self-adaptive energy management and power distribution are achieved based on power requirements, energy recovery is triggered during sudden braking or downhill, and safe operation is ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of data analysis, and in particular to an electric bicycle state monitoring method based on sensor fusion. Background Art

[0002] With the widespread adoption of electric bicycles in urban transportation, shared mobility, and outdoor sports, achieving real-time monitoring of their operating status and intelligent energy management has become a critical technical challenge. Traditional monitoring methods typically rely on data from a single or small amount of sensor data to determine vehicle status, failing to fully reflect complex driving environments and rider behavior. This leads to delayed response and inaccurate predictions in energy management and safety control.

[0003] Existing research uses multi-sensor data fusion to monitor e-bikes. However, most approaches focus solely on physical parameters, such as air resistance, rolling resistance, and the effects of slope, ignoring the dynamic changes in human-machine synergy and the actual energy consumption characteristics of the road surface during riding. Traditional approaches also have limitations in energy recovery and path planning. On the one hand, they fail to fully exploit the synergy between the rider's pedaling input and the motor's output; on the other hand, the impact of road conditions on energy consumption is difficult to quantify, resulting in a lack of scientific basis for energy recovery strategies and path planning decisions. Summary of the Invention

[0004] Based on the above-mentioned shortcomings of the prior art, the purpose of the present invention is to provide an electric bicycle state monitoring method based on sensor fusion to solve the above-mentioned technical problems.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a method for monitoring the state of an electric bicycle based on sensor fusion, comprising: Multi-dimensional data is collected by multi-dimensional sensors, and the resonance synergy index and road energy transparency index are calculated based on the multi-dimensional data, and the global feature vector is obtained by fusion; The global feature vector of the time series is input into the preset dual-branch prediction model to output the riding intention and riding mode; Calculate power requirements based on multi-dimensional data, and adaptively distribute power based on power requirements and riding mode. When the riding intention is braking or going downhill, energy is recovered according to the preset energy recovery strategy. The predicted energy consumption of the road section is calculated based on the power demand, the cost index of the road section is calculated based on the predicted energy consumption and the road energy transparency index, and the path planning is performed based on the predicted energy consumption and the cost index.

[0006] The present invention is further configured such that the multidimensional data includes acceleration, angular velocity, pedal torque, motor output torque, pedal speed, slope angle, vehicle speed and vibration signal variance.

[0007] The present invention is further configured such that the calculation logic of the resonance synergy index is: , for The resonance synergy index of the moment, is the time window, and for The instantaneous phase of the pedal torque and the motor output torque at the moment; The calculation logic of the road energy transparency index is: , for The road transparency index at each moment, is the slope angle, for The variance of the vibration signal at time for The speed of the car at the moment, is the rolling resistance coefficient, is the friction decay rate, is a constant; The multidimensional data, resonance synergy index and road energy transparency index are fused in any order to obtain the global feature vector.

[0008] The present invention is further configured to input the global feature vector under the time series into a preset dual-branch prediction model, and perform riding intention prediction through a neural network including a long short-term memory network; Introducing an attention gating mechanism based on resonance cooperativity index in long short-term memory networks; Based on the attention gate output, the Softmax function is used to generate the probability distribution of riding intention; The global feature vector under the time series is input into the preset Gaussian mixture model to cluster the riding patterns, and the pattern confidence is output by modifying the confidence calculation logic.

[0009] The present invention is further configured such that the gating function is: , is the gating function, Long short-term memory network The hidden state of the moment, for The resonance synergy index of the moment, and is the weight matrix, is the Sigmoid activation function; The calculation logic of the probability distribution of cycling intention is: , For a given global eigenvector Under the condition of The predicted probability distribution of is the weight matrix of the output layer, is the normalization function; The calculation logic of pattern confidence is: , For a given global eigenvector Under the conditions, riding mode The predicted probability distribution of is the number of Gaussian distributions used in the Gaussian mixture model, For the The mixture weights of Gaussian components, For the Gaussian components, calculate the global eigenvector The probability density of is the mean vector, is the covariance matrix, is the road condition attenuation coefficient, for The road transparency index at all times.

[0010] The present invention is further configured to calculate the resistance mapping index and the riding input power based on the multi-dimensional data; Calculate total power requirement based on resistance mapping index and riding input power; Adjust the motor power limit according to riding mode, resonance synergy index and baseline power; The actual output power is determined based on the motor power upper limit and the total required power, and adaptive power distribution is performed.

[0011] The present invention is further configured such that the calculation logic of the resistance mapping index is: , is the resistance mapping index, is the air density, is the air resistance coefficient, is the frontal windward area, is the vehicle speed, is the total mass, is the acceleration due to gravity, is the road energy transparency index, is the rolling resistance coefficient, is the slope angle; The calculation logic of cycling input power is: , Input power for cycling, is the pedal torque, is the pedal speed; The calculation logic of the total required power is: , is the total required power, is the efficiency factor, ; The calculation logic of the motor power upper limit is: , is the upper limit of motor power, is the riding mode coefficient, is the reference power, is the synergistic adjustment coefficient, is the resonance synergy index; The calculation logic of the actual output power is: , is the actual output power.

[0012] The present invention is further configured to calculate the energy recovery ratio based on the road energy transparency index when the riding intention is braking or on a downhill slope; The recovery power is calculated according to the energy recovery ratio and energy recovery is performed.

[0013] The present invention is further configured such that the calculation logic of the energy recovery ratio is: , is the energy recovery ratio, is the road energy transparency index, is the maximum state of charge of the battery, is the current state of charge of the battery; The calculation logic of the recovered power is: , To recover power, is the energy recovery efficiency, is the resistance mapping index, is the vehicle speed, The maximum power limit of the charging system.

[0014] The present invention is further configured such that the calculation logic of the predicted energy consumption of a road section is: , For road sections The predicted energy consumption, is the total required power, is the penalty coefficient, is the road energy transparency index, is the vehicle speed; The calculation logic of the cost index of a road section is: , For road sections The cost index, is the maximum allowable energy consumption, is the energy consumption weight coefficient, is the road condition weight coefficient; Under the constraints, the path planning is performed by minimizing the predicted energy consumption and cost index of the road section.

[0015] The present invention provides an electric bicycle state monitoring method based on sensor fusion. The method collects multidimensional data through multidimensional sensors, calculates a resonance synergy index and a road energy transparency index based on the multidimensional data, and fuses them to obtain a global feature vector. The global feature vector in a time series is input into a preset dual-branch prediction model to output riding intention and riding mode. The method calculates power demand based on the multidimensional data, performs adaptive power allocation based on the power demand and riding mode, and performs energy recovery according to a preset energy recovery strategy when the riding intention is braking or going downhill. The method calculates the predicted energy consumption of a road section based on the power demand, calculates the cost index of the road section based on the predicted energy consumption and the road energy transparency index, and performs path planning based on the predicted energy consumption and the cost index. The beneficial effects produced include: 1. Comprehensive Perception and Real-Time Monitoring: This system collects data through multi-dimensional sensors and quantifies the synergy between rider and motor, as well as road conditions, using the Resonance Synergy Index and Road Energy Transparency Index. This fully reflects the operating status of the e-bike, effectively resolving the problem that traditional single-sensor monitoring cannot capture complex environments and human-machine collaboration. 2. Accurate Cycling Intention and Pattern Prediction: A dual-branch prediction model uses an attention gating mechanism based on a long short-term memory network and a Gaussian mixture model to cluster global features. This allows for accurate prediction of cycling intentions and patterns over time, providing a reliable basis for subsequent energy management and route planning. 3. Adaptive Energy Management and Power Allocation: Based on power demand calculated from multi-dimensional data, combined with riding mode and RSI information, adaptive power allocation is achieved by dynamically adjusting the motor power limit. This method reduces energy consumption while meeting driving requirements, fully utilizing the rider's input power, improving overall system energy utilization, and ensuring timely triggering of energy recovery strategies during sudden braking or downhill conditions, protecting system safety.

[0016] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without inventive efforts. In the drawings: Figure 1The flowchart of a method for monitoring the state of an electric bicycle based on sensor fusion is shown as an exemplary embodiment of the present invention. DETAILED DESCRIPTION

[0018] The following describes the embodiments of the present invention with reference to the accompanying drawings and preferred embodiments. Those skilled in the art will readily appreciate the other advantages and benefits of the present invention from the disclosure herein. The present invention may also be implemented or applied through various other specific embodiments, and the various details in this specification may be modified or altered based on different viewpoints and applications without departing from the spirit of the present invention. It should be understood that the preferred embodiments are intended only to illustrate the present invention and are not intended to limit the scope of protection of the present invention.

[0019] It should be noted that the illustrations provided in the following embodiments are merely schematic illustrations of the basic concept of the present invention. Therefore, the illustrations only show components related to the present invention and are not drawn according to the number, shape, and size of components in actual implementation. In actual implementation, the type, quantity, and proportion of each component may be changed arbitrarily, and the component layout may also be more complex.

[0020] In the following description, numerous details are discussed to provide a more thorough explanation of the embodiments of the present invention. However, it will be apparent to those skilled in the art that the embodiments of the present invention may be practiced without these specific details. In other embodiments, well-known structures and devices are shown in block diagram form rather than in detail to avoid obscuring the embodiments of the present invention.

[0021] A state monitoring method for electric bicycles based on sensor fusion, such as Figure 1 Shown, including: Multi-dimensional data is collected by multi-dimensional sensors, and the resonance synergy index and road energy transparency index are calculated based on the multi-dimensional data, and the global feature vector is obtained by fusion; The global feature vector of the time series is input into the preset dual-branch prediction model to output the riding intention and riding mode; Calculate power requirements based on multi-dimensional data, and adaptively distribute power based on power requirements and riding mode. When the riding intention is braking or going downhill, energy is recovered according to the preset energy recovery strategy. The predicted energy consumption of the road section is calculated based on the power demand, the cost index of the road section is calculated based on the predicted energy consumption and the road energy transparency index, and the path planning is performed based on the predicted energy consumption and the cost index.

[0022] The present invention further provides that the multidimensional data includes acceleration, angular velocity, pedal torque, motor output torque, pedal speed, slope angle, vehicle speed, and vibration signal variance. Specifically, acceleration represents the rate of change of linear motion of the electric bicycle and its rider, reflecting the acceleration, deceleration, and steering conditions of the vehicle at a given moment. This is acquired via an inertial measurement unit (IMU) mounted on the vehicle. The IMU includes a built-in three-axis accelerometer that can collect real-time acceleration data in the horizontal (x, y) and vertical (z) directions. After preprocessing (such as filtering and gravity compensation), this data is used to construct a vehicle acceleration vector, reflecting the vehicle's dynamic motion state. Angular velocity describes the vehicle's rotational speed about a specific axis (typically the vertical or horizontal axis), reflecting cornering, tilting, or bumping. This is also acquired by the IMU's built-in gyroscope. The gyroscope provides angular velocity data about three orthogonal axes, which is used to determine the vehicle's rotational motion. After correction and filtering, this data is used to assist in vehicle posture estimation and motion state analysis. Pedal torque refers to the rotational torque applied by the rider by pedaling, directly reflecting the rider's physical output and effort. It is acquired via a torque sensor installed on the pedals or drivetrain. The torque sensor converts mechanical torque into an electrical signal, which is then digitized by a data acquisition module to generate real-time pedal torque data. Motor output torque represents the amount of power applied by the motor to the wheels, reflecting the power output of motor-assisted driving. It is measured by a torque sensor built into or connected to the motor controller. The actual output torque is calculated by monitoring internal motor current, voltage, and other signals, combined with the motor's characteristic curve. Pedal speed refers to the rotational speed of the pedals, typically expressed in revolutions per minute (RPM) or radians per second. This reflects the rider's pedaling frequency. Pedal rotation is detected by a speed sensor (such as an encoder or Hall effect sensor) mounted on the pedal shaft or related components. This is converted into a digital signal to generate real-time speed data. The slope angle represents the current road's inclination relative to the horizontal, reflecting the degree of uphill or downhill travel and directly affecting the gravity component and energy consumption. It is calculated using an incline sensor or accelerometer data from the IMU (combined with the direction of gravity). Some systems may integrate a dedicated slope sensor or use GPS and map data to aid in estimating road slope. Vehicle speed is the actual speed of the vehicle while in motion and is a key parameter affecting air resistance, acceleration requirements, and energy consumption. Vehicle speed can be obtained from a variety of sensors, such as a GPS module (which calculates speed based on position changes) or a wheel speed sensor (which calculates linear speed based on wheel diameter). Multi-sensor data fusion can improve the accuracy and robustness of vehicle speed measurement; the vibration signal variance reflects the vibration intensity caused by uneven road surface, bumps and other factors during vehicle driving.A high variance value typically indicates a rough road surface or significant impact on the vehicle, potentially affecting energy transfer efficiency. High-frequency vibration sensors (or accelerometers) installed in key vehicle locations (such as the frame or chassis) collect vibration data. This data is statistically processed (for example, by calculating variance within a specific time window) to determine the variance of the vibration signal. After filtering and normalization, this data can be used as an important indicator of road condition. After acquiring this data through multi-dimensional sensors, it is uploaded to the cloud via a communication module for analysis.

[0023] The present invention is further configured such that the calculation logic of the resonance synergy index is: , for The resonance synergy index of the moment, is the time window, and for The instantaneous phase of the pedal torque and the motor output torque at the moment; specifically, the instantaneous phase difference between the pedal torque signal and the motor output torque signal is calculated to quantify the synergy between the rider's input and the motor assistance. The better the synergy, the more synchronized the phase between the rider's input and the motor output; the Hilbert transform is used to convert the original torque signal into an analytical signal, thereby extracting the instantaneous phase. , , and Respectively indicate time Pedal torque and motor output torque, is the Hilbert transform, which is used to convert the real signal into an analytical signal to obtain the instantaneous phase of the signal. Indicates calculating the complex angle, that is, the instantaneous phase; calculating the difference between the instantaneous phases of the two signals, and then taking the cosine to obtain a real-time indicator reflecting the degree of synchronization. Integrate and average the cosine value to get the time The resonance synergy index provides a quantitative indicator for real-time evaluation of the phase synchronization between the rider's input and the motor's output, enabling the system to dynamically adjust power distribution and energy recovery strategies based on the actual synergy effect. By monitoring the resonance synergy index, the system can better balance the rider's power input and motor output, ensuring full utilization of the rider's physical input under good synergy, thereby reducing the burden on the motor and optimizing overall energy consumption.

[0024] The calculation logic of the road energy transparency index is: , for The road transparency index at each moment, is the slope angle, for The variance of the vibration signal at time for The speed of the car at the moment, is the rolling resistance coefficient, is the friction decay rate, is a constant; specifically, the road energy transparency index is used to quantify the impact of the road on energy transfer and dissipation, that is, to evaluate the impact of road conditions on vehicle energy consumption. The index comprehensively considers the slope effect, road roughness (reflected by vibration signals), and the attenuation of rolling resistance over time or road conditions. Indicates the effect of road slope on energy consumption. This item increases energy consumption when going uphill, and the opposite is true when going downhill. Reflects the degree of road roughness, divided by vehicle speed Normalization is used to ensure that the impact of road roughness on energy consumption is more significant at low speeds. Represents rolling resistance and friction attenuation effect, describing the change of road friction effect with time or state. Adding the above three parts, we can get Road energy transparency index; The multidimensional data, resonance synergy index and road energy transparency index are fused in any order to obtain the global feature vector.

[0025] The present invention is further configured to input the global feature vector under the time series into a preset dual-branch prediction model, and perform riding intention prediction through a neural network including a long short-term memory network; An attention gating mechanism based on the resonance synergy index is introduced into the long short-term memory network; the present invention is further configured such that the gating function is: , is the gating function, Long short-term memory network The hidden state of the moment, for The resonance synergy index of the moment, and is the weight matrix, is the Sigmoid activation function; Based on the attention gate output, the Softmax function is used to generate the probability distribution of riding intention. The calculation logic of the probability distribution of riding intention is as follows: , For a given global eigenvector Under the condition of The predicted probability distribution of is the weight matrix of the output layer, is the normalization function; The global feature vectors in the time series are input into the preset Gaussian mixture model to cluster the riding patterns. The pattern confidence is then output by modifying the confidence calculation logic. The calculation logic of the pattern confidence is as follows: , For a given global eigenvector Under the conditions, riding mode The predicted probability distribution of is the number of Gaussian distributions used in the Gaussian mixture model, For the The mixture weights of Gaussian components, For the Gaussian components, calculate the global eigenvector The probability density of is the mean vector, is the covariance matrix, is the road condition attenuation coefficient, for The road transparency index at all times.

[0026] The present invention is further configured to calculate the resistance mapping index and the riding input power based on the multidimensional data; the present invention is further configured to calculate the resistance mapping index using the following logic: , is the resistance mapping index, is the air density, is the air resistance coefficient, is the frontal windward area, is the vehicle speed, is the total mass, is the acceleration due to gravity, is the road energy transparency index, is the rolling resistance coefficient, is the slope angle; specifically, the resistance mapping index Comprehensively reflects the air resistance, slope effect and road conditions through the road energy transparency index and rolling resistance coefficient Impact on vehicle energy consumption; Item 1 represents air resistance, which changes with the quadratic power of vehicle speed; the second term The total drag mapping index is calculated by combining the effects of gravity, road conditions, and rolling resistance on energy consumption.

[0027] The calculation logic of cycling input power is: , Input power for cycling, is the pedal torque, is the pedal speed; specifically, the riding input power Represents the mechanical energy provided by the rider. As part of the vehicle's power, this power value reflects human input and is an important reference for determining human-machine collaboration and adjusting motor output in the energy management system. The total required power is calculated based on the resistance mapping index and the riding input power. The calculation logic of the total required power is: , is the total required power, is the efficiency factor, Total power demand The calculation logic is composed of three parts: Resistance overcoming part: This part represents the vehicle's current driving state in order to overcome air resistance, gravity, rolling resistance and other road dissipation factors (measured by the resistance mapping index The power required is expressed as It is a linear relationship; acceleration power part: , which represents the additional power required to accelerate the vehicle, based on the vehicle's mass ,speed and acceleration The calculation reflects the energy consumption corresponding to the rate of change of vehicle kinetic energy; the rider's input contribution: , which takes into account the mechanical energy provided by the rider by pedaling. By introducing the efficiency factor , by the resonance synergy index Modulation, calculated as , only when When it exceeds 0.5, the rider's input can be effectively converted into energy available to the system, thereby reducing the need for supplementary motors; taking into account the energy required for the vehicle to overcome resistance and accelerate, as well as the rider's actual power input, the total required power Can accurately reflect actual energy demand and avoid insufficient motor output or energy waste; The motor power limit is adjusted based on the riding mode, resonance synergy index, and baseline power. The calculation logic for the motor power limit is: , is the upper limit of motor power, is the riding mode coefficient, is the reference power, is the synergistic adjustment coefficient, is the resonance synergy index; specifically, in the energy management of electric bicycles, in order to ensure that the motor output meets the vehicle power requirements while not exceeding the safety and design limits, a dynamic motor power upper limit needs to be set . This upper limit should take into account the different energy consumption requirements of the current riding mode, which includes commuting, exercise and load-bearing, as well as the human-machine synergy effect, that is, the synergy between the rider and the motor. The resonance synergy index is an indicator to measure the effectiveness of human-machine synergy. When the resonance synergy index is high, it means that the synchronization between the rider's input and the motor's output is good. At this time, the system can appropriately increase the motor power upper limit to make full use of the rider's effective input; otherwise, the power upper limit is lowered to prevent energy waste or safety hazards caused by poor synergy. Therefore, the motor power upper limit The calculation logic combines the riding mode coefficient , reference power The synergy factor is adjusted based on the resonance synergy index to dynamically determine the maximum motor output power allowed in the current state. The motor power limit is dynamically adjusted according to the riding mode and the state of human-machine cooperation, so that the motor can output higher power when the synergy effect is good, thereby more efficiently cooperating with the rider's input, improving power response and energy utilization. The actual output power is determined based on the motor power upper limit and the total required power, and adaptive power distribution is performed. The calculation logic of the actual output power is: , is the actual output power; specifically, in the energy management of electric bicycles, the system needs to be based on the current total power demand and the rider's input power Determine the power added by the motor, subject to the upper limit of the motor's output Total power demand Indicates the total energy required by the vehicle to overcome external resistance, accelerate, etc.; riding input power Reflects the mechanical energy provided by the rider through pedaling; residual power demand This is the part that the motor needs to supplement; in order to prevent the motor output from exceeding the safety and design upper limit, the system combines this residual demand with Compare and take the smaller value as the actual output power ; By taking a smaller value, ensure that the motor output does not exceed the design upper limit , to prevent potential safety hazards caused by overload and make full use of the rider's input power To offset part of the total demand, so that the motor is only responsible for supplementing the shortfall and improving the overall energy utilization.

[0028] The present invention is further configured such that when the riding intention is braking or riding downhill, the energy recovery ratio is calculated based on the road energy transparency index. The present invention is further configured such that the calculation logic of the energy recovery ratio is: , is the energy recovery ratio, is the road energy transparency index, is the maximum state of charge of the battery, The current state of charge of the battery. Specifically, when braking or going downhill, the vehicle will generate a large amount of braking energy due to gravity and deceleration. In order to improve energy utilization efficiency, the system will try to recover this energy. Energy recovery ratio Used to adjust the working intensity of the recovery system. Its calculation logic comprehensively considers: road conditions: determined by the road energy transparency index Characterization. Higher Indicates poor road conditions, including steep slopes, bumps or rough roads, which may limit the energy recovery effect during braking; Battery charging status: The current battery status Maximum state of charge If the battery charge margin is sufficient, a larger proportion of energy recovery is allowed; otherwise, energy recovery needs to be limited to protect the battery. Finally, the two are combined and then The maximum value is limited to 1 to obtain the energy recovery ratio. By comprehensively considering road conditions and battery charge remaining, the energy recovery system can intelligently adjust the recovery ratio during braking or downhill driving to ensure that braking energy is fully recovered when there is sufficient charge remaining, reducing energy waste. The recovery power is calculated based on the energy recovery ratio to perform energy recovery. The calculation logic of the recovery power is as follows: , To recover power, is the energy recovery efficiency, is the resistance mapping index, is the vehicle speed, The maximum power limit of the charging system. Specifically, when braking or going downhill, the braking energy generated by the vehicle can be recovered and converted into electrical energy and stored in the battery. To this end, the system needs to determine an appropriate recovery power : Map index based on the vehicle's current resistance (reflecting the external resistance of the vehicle) and vehicle speed , calculate the available energy of the vehicle during braking, that is, .here, It is the energy recovery ratio calculated based on the road conditions and battery charge state, reflecting the energy recovery ratio allowed under the current conditions. In order to prevent the charging system from exceeding the safety range, the system compares the calculated value with the maximum power limit of the charging system. Compare the two and take the smaller value. Finally, multiply by the energy recovery efficiency , and obtain the power that can actually be recovered and converted into electrical energy; by combining road conditions and battery status, the system can dynamically calculate the recovery ratio to ensure that energy is fully recovered when there is sufficient charging margin and suitable road conditions, thereby improving energy utilization efficiency.

[0029] The present invention is further configured such that the calculation logic of the predicted energy consumption of a road section is: , For road sections The predicted energy consumption, is the total required power, is the penalty coefficient, is the road energy transparency index, is the vehicle speed; specifically, Indicates the total power required for the vehicle to overcome resistance and accelerate in its current driving state; is an additional energy penalty term used to reflect the poor road conditions (caused by Indicated by) and the additional energy consumption caused by the speed factor. The integration process is The total predicted energy consumption of the road section is obtained by adding up the values.

[0030] The calculation logic of the cost index of a road section is: , For road sections The cost index, is the maximum allowable energy consumption, is the energy consumption weight coefficient, is the road condition weight coefficient; specifically, The normalized predicted energy consumption reflects the energy consumption of the road section within the allowable energy consumption range. The proportion of energy consumption, energy consumption weight coefficient Used to adjust the impact of energy consumption on cost, Indicates the impact of road conditions on the cost, where is the road energy transparency index of the road section, is the road condition weight coefficient, and finally, As a comprehensive indicator, it is used to compare and select different road sections during route planning; Under the constraints, the predicted energy consumption and cost index of each road section are minimized for path planning. Specifically, in path planning, a comprehensive objective function is constructed, which comprehensively considers the predicted energy consumption of each road section. and price index The planning algorithm selects the path that minimizes this combined objective function, subject to constraints such as remaining energy, travel time, and vehicle safety. This is a state-of-the-art approach and will not be detailed here. The final path planning solution considers the energy consumption and road surface cost of each road segment while also satisfying energy, time, and safety constraints, thereby achieving energy optimization and safe driving for the entire system.

[0031] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer program are loaded or executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website, computer, server or data center to another website, computer, server or data center via a wired (e.g., infrared, wireless, microwave, etc.) method. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state drive.

[0032] It should be understood that the term "and / or" as used herein simply describes a relationship between associated objects, indicating that three possible relationships exist. For example, "A and / or B" can represent: A alone, A and B together, or B alone. A and B can be singular or plural. Furthermore, the character " / " as used herein generally indicates an "or" relationship between the associated objects, but it may also indicate an "and / or" relationship. For specific understanding, please refer to the context.

[0033] In this application, "at least one" means one or more, and "plurality" means two or more. "At least one of the following" or similar expressions refers to any combination of these items, including any combination of single or plural items. For example, at least one of a, b, or c can mean: a, b, c, ab, ac, bc, or abc, where a, b, and c can be single or plural.

[0034] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0035] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0036] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0037] In the several embodiments provided in this application, it should be understood that the disclosed system can be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of the units is merely a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0038] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0039] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0040] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0041] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

Claims

1. A method for monitoring the state of an electric bicycle based on sensor fusion, characterized in that: include: Multi-dimensional data is collected by multi-dimensional sensors, and the resonance synergy index and road energy transparency index are calculated based on the multi-dimensional data, and the global feature vector is obtained by fusion; The global feature vector of the time series is input into the preset dual-branch prediction model to output the riding intention and riding mode; Calculate power requirements based on multi-dimensional data, and adaptively distribute power based on power requirements and riding mode. When the riding intention is braking or going downhill, energy is recovered according to the preset energy recovery strategy. The predicted energy consumption of the road section is calculated based on the power demand, the cost index of the road section is calculated based on the predicted energy consumption and the road energy transparency index, and the path planning is performed based on the predicted energy consumption and the cost index.

2. The electric bicycle state monitoring method based on sensor fusion according to claim 1 is characterized in that: The multidimensional data includes acceleration, angular velocity, pedal torque, motor output torque, pedal speed, slope angle, vehicle speed and vibration signal variance.

3. The electric bicycle state monitoring method based on sensor fusion according to claim 2 is characterized in that: The calculation logic of the resonance synergy index is: , for The resonance synergy index of the moment, is the time window, and for The instantaneous phase of the pedal torque and the motor output torque at the moment; The calculation logic of the road energy transparency index is: , for The road transparency index at each moment, is the slope angle, for The variance of the vibration signal at time for The speed of the car at the moment, is the rolling resistance coefficient, is the friction decay rate, is a constant; The multidimensional data, resonance synergy index and road energy transparency index are fused in any order to obtain the global feature vector.

4. The electric bicycle state monitoring method based on sensor fusion according to claim 1 is characterized in that: The global feature vector under the time series is input into the preset dual-branch prediction model, and the riding intention is predicted through the neural network including the long short-term memory network; Introducing an attention gating mechanism based on resonance cooperativity index in long short-term memory networks; Based on the attention gate output, the Softmax function is used to generate the probability distribution of riding intention; The global feature vector under the time series is input into the preset Gaussian mixture model to cluster the riding patterns, and the pattern confidence is output by modifying the confidence calculation logic.

5. The electric bicycle state monitoring method based on sensor fusion according to claim 4 is characterized in that: The gating function is: , is the gating function, Long short-term memory network The hidden state of the moment, for The resonance synergy index of the moment, and is the weight matrix, is the Sigmoid activation function; The calculation logic of the probability distribution of cycling intention is: , For a given global eigenvector Under the condition of The predicted probability distribution of is the weight matrix of the output layer, is the normalization function; The calculation logic of pattern confidence is: , For a given global eigenvector Under the conditions, riding mode The predicted probability distribution of is the number of Gaussian distributions used in the Gaussian mixture model, For the The mixture weights of Gaussian components, For the Gaussian components, calculate the global eigenvector The probability density of is the mean vector, is the covariance matrix, is the road condition attenuation coefficient, for The road transparency index at all times.

6. The electric bicycle state monitoring method based on sensor fusion according to claim 1 is characterized in that: Calculate resistance mapping index and cycling input power based on multi-dimensional data; Calculate total power requirement based on resistance mapping index and riding input power; Adjust the motor power limit according to riding mode, resonance synergy index and baseline power; The actual output power is determined based on the motor power upper limit and the total required power, and adaptive power distribution is performed.

7. The electric bicycle state monitoring method based on sensor fusion according to claim 6 is characterized in that: The calculation logic of the resistance mapping index is: , is the resistance mapping index, is the air density, is the air resistance coefficient, is the frontal windward area, is the vehicle speed, is the total mass, is the acceleration due to gravity, is the road energy transparency index, is the rolling resistance coefficient, is the slope angle; The calculation logic of cycling input power is: , Input power for cycling, is the pedal torque, is the pedal speed; The calculation logic of the total required power is: , is the total required power, is the efficiency factor, ; The calculation logic of the motor power upper limit is: , is the upper limit of motor power, is the riding mode coefficient, is the reference power, is the synergistic adjustment coefficient, is the resonance synergy index; The calculation logic of the actual output power is: , is the actual output power.

8. The electric bicycle state monitoring method based on sensor fusion according to claim 1 is characterized in that: When the riding intention is to brake or ride downhill, the energy recovery ratio is calculated based on the road energy transparency index; The recovery power is calculated according to the energy recovery ratio and energy recovery is performed.

9. The electric bicycle state monitoring method based on sensor fusion according to claim 8, characterized in that: The calculation logic of energy recovery ratio is: , is the energy recovery ratio, is the road energy transparency index, is the maximum state of charge of the battery, is the current state of charge of the battery; The calculation logic of the recovered power is: , To recover power, is the energy recovery efficiency, is the resistance mapping index, is the vehicle speed, The maximum power limit of the charging system.

10. The electric bicycle state monitoring method based on sensor fusion according to claim 1, characterized in that: The calculation logic of the predicted energy consumption of a road section is: , For road sections The predicted energy consumption, is the total required power, is the penalty coefficient, is the road energy transparency index, is the vehicle speed; The calculation logic of the cost index of a road section is: , For road sections The cost index, is the maximum allowable energy consumption, is the energy consumption weight coefficient, is the road condition weight coefficient; Under the constraints, the path planning is performed by minimizing the predicted energy consumption and cost index of the road section.

Images