CN109131928B - Identification method and device for electric power system of light unmanned aerial vehicle - Google Patents

Abstract

The invention discloses a method and a device for identifying a light unmanned aerial vehicle electric power system, which comprises the following steps: building a Hammerstein nonlinear differential equation structure of a time domain mathematical model of the electric power system of the light unmanned aerial vehicle, wherein the Hammerstein nonlinear differential equation structure comprises a first-order linear differential link and a second-order nonlinear link; identifying linear link parameters: the photoelectric coding discs with equal rotational inertia replace an air propeller, so that the function of simulating the inertia force of accelerated rotation is achieved, and the time constant and the amplification coefficient of a first-order linear differential link can be obtained by giving a step signal and outputting a rotational speed response; identifying nonlinear link parameters: the nonlinear link does not relate to differential terms, and parameters of the nonlinear link can be fit and calculated through multiple groups of steady-state input and output relations. Compared with the traditional technology, the method has higher objectivity and accuracy, and can better avoid the adverse effect caused by the nonlinear term generated by the propeller aerodynamic force during the dynamic identification of the differential term; the invention uses the ground testing device, and can accurately acquire the rotating speed response on line.

Description

Identification method and device for electric power system of light unmanned aerial vehicle

Technical Field

The invention relates to the technical field of light unmanned aerial vehicles, in particular to a method and a device for identifying an electric power system of a light unmanned aerial vehicle.

Background

Light unmanned aerial vehicle in civilian field, especially many rotor unmanned aerial vehicle adopt the electric power device more. The typical electric power device consists of an electronic speed regulator, a direct current motor and an air propeller, converts electric energy from a storage battery into aerodynamic force, and is the main power for providing lift force or posture adjustment for the unmanned aerial vehicle. The aerodynamic torque is monotonically related to the rotating speed (or shaft angular speed) of the air propeller, and the rotating speed is determined by the direct current power regulated by the digital signal. Therefore, the time domain relation between the digital signal and the rotating speed of the air propeller forms an input-output model of the unmanned aerial vehicle electric power device, and the accuracy of the model has a crucial influence on the simulation design of a controller of the electric power device. The input digital signal can be directly collected by a single chip microcomputer, and the output rotating speed is obtained by a photoelectric sensor through collecting the time of a code disc or a propeller blade sweeping photoelectric gate.

The input of the electronic Speed regulator (Electric Speed Controller) is a digital signal representing the degree of 0-100% (or high level time proportion) and a constant voltage DC power supply, and the output DC voltage is the product of the percentage represented by the digital signal and the DC power supply pressure.

A direct current Motor (Brushless DC Motor) converts a direct current voltage output from an electronic governor into a rotor torque. For a small direct current motor, if armature inductance and bearing friction resistance are not considered, the magnitude of the rotor no-load rotation speed and direct current voltage are in a linear relation in a steady state, and the rotor no-load rotation speed and the direct current voltage are in a first-order inertia relation in a dynamic state.

An air propeller (Aeropropeller) consumes the torque output by the direct current motor and obtains a certain rotating speed, and the rotating blades interact with air to generate blade surface lift. According to the classic phyllotoxin theory derivation, the real-time consumed torque of the air propeller and the rotating speed form a nonlinear relation.

The photoelectric Sensor (Electro-optical Sensor) consists of a photoelectric gate and a digital signal modulation circuit, when a propeller blade or a code disc sweeps the photoelectric gate to block a light path, the system automatically outputs a high-level signal, otherwise, a low-level signal is output. The single chip microcomputer is converted into the rotating speed of the propeller or the coding disc by calculating the proportion of the high level time to the total time.

Through a theoretical framework and mathematical derivation, the time domain mathematical model of the unmanned aerial vehicle electrodynamic force system inputs pulse width modulation signals (PWM) and outputs the propeller rotating speed (or shaft angular speed), and the time domain mathematical model is a nonlinear equation with a mixed term. In the case that a plurality of hardware characteristic parameters are unknown, a plurality of parameters in the equation can be obtained only through identification, and the identification precision and operability become key factors in engineering practice.

There are many system parameter identification methods, and for linear differential systems, a signal with rich frequency spectrum is mostly input first, and corresponding parameters are obtained by analyzing the characteristic calculation of the output response signal. For nonlinear differential systems, it is difficult to identify parameters by a general method due to the wide variety of systems. Usually, the terms can be distinguished separately if the non-linear differential system consists of only a linear differential element and a non-linear element.

Disclosure of Invention

The invention aims to be as follows: 1) providing a system parameter identification method, and establishing a time domain nonlinear model of a light unmanned aerial vehicle electric power system; 2) the method for replacing the propeller with equal inertia is provided, so that the rotor is prevented from generating aerodynamic force when rotating, and the rotating speed can be easily measured; 3) the utility model provides a ground testing arrangement suitable for light-duty unmanned aerial vehicle electric power system when realizing accurate collection rotational speed, possesses the multiple performance parameter test function.

The technical scheme adopted by the invention is as follows: a light unmanned aerial vehicle electric power system identification method comprises the following steps:

step 1, building a nonlinear differential equation frame: a time domain mathematical model of a light unmanned aerial vehicle electrodynamic force system belongs to one of Hammerstein (Hammerstein) type nonlinear differential equations and is composed of a first-order linear differential link and a second-order nonlinear link;

step 2, identifying linear link parameters: the photoelectric coding discs with equal rotational inertia replace an air propeller, so that the function of nondestructively simulating the inertia force of accelerated rotation is achieved, and the time constant and the amplification coefficient of a first-order linear differential link can be obtained by giving a step signal and responding according to the output rotating speed;

step 3, identifying nonlinear link parameters: the nonlinear link does not relate to differential terms, and parameters of the nonlinear link can be fit and calculated through multiple groups of steady-state input and output relations.

A light unmanned aerial vehicle electric power system identification device, comprising:

photoelectric coding disc: the ABS material disc is made of an ABS material, 6-9 notches are uniformly distributed on the circumference, and a compensation cylindrical block is designed near the center;

a photoelectric sensor: the width of the photoelectric gate is 0.02m, when the optical path of the photoelectric gate is blocked, the signal pin of the sensor outputs high level, otherwise, the signal pin outputs low level;

light-duty unmanned aerial vehicle electric power system ground test platform: the system is composed of an AVR type singlechip minimum system, a bending moment type strain sensor (strain beam), an A/D conversion element, a photoelectric sensor, a current-voltage sensor, a non-contact temperature sensor and a linear power rod, can measure parameters such as tension, reaction torque, current, voltage, rotor temperature, rotating speed and the like, can acquire output response data, and can identify and obtain time domain input and output mathematical model parameters of an electrodynamic force system.

The working process is as follows:

the method comprises the following steps: simplifying blades of an air propeller into a rectangle, calculating the rotational inertia of the propeller, reversely designing a disc-shaped coding disc, manufacturing the coding disc by using a material increase technology printer, setting the printing resolution to be maximum, and taking ABS as a material;

step two: installing a ground testing device and a light unmanned aerial vehicle electric power system, using a photoelectric coding disc to replace a propeller, starting a motor by the system, and stabilizing the rotating speed of the coding disc at the lowest self-sustaining rotating speed;

step three: inputting a step PWM signal value for an electrodynamic force system, wherein the upper limit of the step is a signal value corresponding to the maximum power, simultaneously sensing and recording a rotating speed increasing process by using a photoelectric sensor, determining a rotating speed increment according to time domain rotating speed response, finding a time increment corresponding to 63.2% of the rotating speed increment on a curve, wherein the time increment is a time constant of a first-order inertia link, and the ratio of the rotating speed increment to the signal value increment is a proportionality constant of the first-order inertia link;

step four: the coding disc is disassembled and the air propeller is reassembled;

step five: inputting a plurality of constant PWM signal values for an electrodynamic force system, sensing and recording the rotating speed by using a photoelectric sensor to obtain a plurality of stable rotating speed outputs, and fitting and calculating a parameter value of a quadratic power term by using a plurality of groups of input and output data;

step six: and (5) obtaining a time domain input and output mathematical model of the electric power system of the light unmanned aerial vehicle through sorting, and finishing the identification process.

The working principle of the invention is as follows: A) and a nonlinear differential equation mathematical model: in an electronic governor, the value of an input signal (the value of a pulse width modulated signal, i.e., a PWM value) is proportional to the output voltage. When the direct current motor is in operation, a dynamic relation between an input voltage and an output load torque (propeller consumption torque) is derived from a relation that the sum of an induced potential, a winding potential and an induced potential is equal to an input potential and a relation that the sum of a shaft load torque, a damping torque and an inertia torque is equal to an armature current generation torque. The relationship between the output rotating speed and the input torque of the air propeller is obtained by the air propeller phyllotactic theory. Therefore, three input-output relational expressions are combined, and the input-output time domain mathematical model of the electrodynamic force system is deduced to be a Hammerstein nonlinear differential equation after slight simplification in consideration of the small armature inductance of the mild motor. B) And the rotational inertia photoelectric encoding disk: firstly, a single blade of the air propeller is simplified into a rectangular blade which is provided with the length of a blade expansion, the width of the blade expansion, the thickness of the blade expansion and the thickness of the blade expansion, and is not twisted, and the rotational inertia of the propeller is equal to the sum of the rotational inertia of all the blades. And then, reversely designing a disk-shaped coding disk with moderate thickness according to the principle that the rotational inertia is equal, wherein the number, the width and the depth of the gaps are determined by the size of the photoelectric gate, and the rotational inertia reduced by the gaps is offset with the rotational inertia increased by the compensation cylinder. C) The sampling frequency of the singlechip and the photoelectric sensor is set to be 500 times/second in consideration of the short acceleration time of the photoelectric coding disc or the propeller under the condition that the motor outputs full power. The brushless direct current motor has a dead zone when starting, so the initial rotating speed before inputting step signals to carry out full-power acceleration is the lowest self-sustaining rotating speed of the motor. D) The nonlinear differential equation (Hammerstein type differential equation) for describing the power system of the light unmanned aerial vehicle consists of a first-order linear link (inertia link) and a second-power nonlinear term. In order to avoid the difficulty of directly identifying the parameters of the nonlinear differential equation, a special coding disc is used for replacing an air propeller, and nonlinear load brought by aerodynamics is avoided, so that the parameters of an inertia link are identified by dynamic input signal values. Then, the propeller is installed again, and the parameters of the second power terms are identified according to the steady-state input signal value. Since the second power term and the inertia link are linearly superposed, the step-by-step identification can complete the identification of all parameters of the nonlinear differential equation.

The invention has the beneficial effects that:

(1) the time domain input and output mathematical model framework of the electric power system of the light unmanned aerial vehicle is established by adopting a component method, and the method has higher objectivity and accuracy compared with a black box identification method;

(2) by adopting an equal-rotational-inertia replacement method, the coding disc has the same rotational inertia as that of the air propeller, so that adverse effects caused by nonlinear terms generated by propeller aerodynamic force can be better avoided during differential term dynamic identification;

(3) by using the ground testing device, the lateral movement of the lower washing air flow can greatly reduce the influence of the ground effect on the pneumatic load, and the rotating speed response can be accurately acquired on line.

Drawings

FIG. 1 is a schematic diagram of a design concept of a code wheel.

Fig. 2 is a schematic diagram of the ground testing apparatus.

Fig. 3 is a schematic diagram of the operation of the electric power system with the air propeller installed.

FIG. 4 is a schematic diagram of the operation of the system for mounting the code wheel.

Fig. 5 is a time domain mathematical model and a composition diagram of the electric power system of the light unmanned aerial vehicle.

Fig. 6 is a schematic view of the air propellers composing the electric power system of the light unmanned aerial vehicle.

Fig. 7 is a functional schematic diagram of the ground device.

Fig. 8 is a comparison graph of the simulated rotational speed output of the mathematical model and the rotational speed output of the physical system after the same step PWM signal value is input.

FIG. 9 is a comparison graph of the mathematical model simulation rotational speed output and the physical system rotational speed output after the same sinusoidal PWM signal value is input.

Detailed Description

The invention is further described with reference to the following figures and detailed description.

FIG. 1 is a schematic diagram of a design concept of a code wheel. Firstly, enabling an air propeller blade in an electrodynamic force system of a light unmanned aerial vehicle to be equivalent to a non-torsional rectangular blade, wherein the length is the span length of the blade, the width is the average chord length of the blade, and the thickness is the thickness at the average chord length, and calculating the rotational inertia of the blade; secondly, the rotational inertia of the coding disc is equal to that of the air propeller, uniformly distributed gaps are compensated equivalently, and the width of each gap is approximately equal to the average chord length of each blade; and finally, manufacturing the coding disc by using an additive technology printer and taking ABS as a material.

Fig. 2 is a schematic diagram of the ground testing apparatus. The device is used for measuring relevant performance parameters of a power system of the light unmanned aerial vehicle, adopts a portable simplified design, can be fixed on a common desktop by a G-shaped clamp, is provided with a torque calibration hole by a cross-shaped sheet metal part provided with a motor, and can be provided with a fixed torque calibration device. The rotating speed or the shaft angular speed is calculated according to the opening and closing time of the photoelectric door by sweeping the U-shaped photoelectric door of the photoelectric sensor at the edge of the propeller blade or the coding disc.

Fig. 3 is a schematic diagram of the operation of the electric power system with the air propeller installed. The AVR singlechip inputs a constant or time-varying PWM signal for the electronic speed regulator, the electronic speed regulator modulates a constant direct current power supply by using a PWM signal value and outputs direct current, and the output voltage and the input PWM signal value are in a direct proportion relation. The DC motor converts an input DC voltage into an output torque, and the output torque is consumed by damping torque, acceleration consumption torque and pneumatic torque together, wherein the damping torque is in direct proportion to the rotating speed, the acceleration consumption torque is in direct proportion to the differential of the rotating speed, and the pneumatic torque is in direct proportion to the second power of the rotating speed. The rotating speed of the propeller is collected by the photoelectric sensor and input to the AVR singlechip.

FIG. 4 is a schematic diagram of the operation of the system for mounting the code wheel. The code wheel only consumes the acceleration torque brought by inertia, and the whole system is equal to an electrodynamic system without pneumatic action force.

Fig. 5 is a time domain mathematical model and a composition diagram of the electric power system of the light unmanned aerial vehicle. The mathematical model of the electrodynamic force system is a Hammerstein nonlinear differential equation, the left side of the equation is a relational expression of output torque, and the right side of the equation is an input PWM signal value. The first-order linear differential link in the red solid line box represents the rotating speed relation of the output of the electronic speed regulator-direct current motor-coding disc combination without air action force, and the second-order polynomial in the blue dotted line box represents the rotating speed relation of the output of the electronic speed regulator-direct current motor-air propeller combination under the condition of steady-state input.

Fig. 6 is a schematic view of the air propellers composing the electric power system of the light unmanned aerial vehicle. From the characteristics of the small air propellers, according to the principles of the folacin and assuming: A) the airflow acts on the surface of the blade, and the airflow of each leaf element is two-dimensional; B) the mutual influence among the phyllanthus and the blades is not counted; C) the washing effect generated by propeller blades is not counted; D) the propeller is rigid and the blades and blade roots are not elastically deformed. The derivation is that the aerodynamic torque is proportional to the second power of the rotational speed.

Fig. 7 is a functional schematic diagram of the ground device. The ground testing device uses a voltage sensor, a current sensor, a photoelectric sensor and a non-contact temperature sensor to measure input voltage, input current, rotating speed and rotor temperature, and uses three bending moment type strain beams to measure tension and reaction torque. After the built-in processor is used for calculation, high-order performance parameters such as input power, consumed electric quantity, propeller power efficiency and the like can be monitored.

Fig. 8 and 9 are graphs comparing the simulation rotational speed output of the mathematical model with the rotational speed output of the physical system. The thick black solid line represents the rotational speed output of a physical system using a 12V-1400kV-2820 model dc motor and an 8 x 4.7 inch air propeller. The thin solid curve represents the simulated rotating speed output of the nonlinear differential equation (graphic equation) established and identified by the invention, and the black dotted line and the gray dotted line represent the simulated rotating speed output of the mathematical model obtained by identifying the physical system by using the traditional first-order inertia link and second-order oscillation link as model frames respectively. The same step PWM signal value (figure 8) and sine PWM signal value (figure 9) are input for a real object system and three different mathematical models, and the coincidence degree of the simulation output of the nonlinear model established by the invention and the rotating speed response of the real object system is found to be highest by comparing rotating speed response curves.

(1) Software and hardware preparation

1) Software aspect:

the software part of the identification method of the electric power system of the light unmanned aerial vehicle mainly comprises a signal output part, a sensing and monitoring part, a high-order performance parameter calculation part and a parameter identification part. The signal output part comprises an AVR singlechip timer and a PWM signal output protocol and is responsible for outputting pulse width modulation signals which can be identified by the electronic speed regulator; the sensing and monitoring part comprises a communication protocol of each sensor, an A/D conversion element, a filter and a resolver, and is responsible for converting analog signals from each sensor into digital signals and processing the digital signals to obtain decimal basic performance parameters with dimensions; the high-order performance parameter calculation part processes part of basic performance parameters again to obtain high-order performance parameters; the parameter identification part is responsible for carrying out mathematical model parameter identification on an electrodynamic force system installed on the ground testing device and outputting various parameters of the mathematical model.

2) Hardware aspect:

the hardware part of the identification device of the electric power system of the light unmanned aerial vehicle mainly comprises a coding disc and a ground testing device. The coding disc is manufactured by a material increase technology printer, and the design principle is that the rotational inertia is equal to that of an air propeller of an identified power system. The ground testing device consists of a G-shaped clamping type fixing device, a sheet metal structural part, a bending moment type strain sensor (strain beam), a Hall current sensor, a mutual inductance type voltage sensor, a non-contact infrared type temperature sensor, a photoelectric door type sensor and a control monitoring terminal. The control monitor terminal is provided with an AVR single chip microcomputer, a liquid crystal display screen and an indicator light.

(2) Operation process

A. Installing the electric power system of the light unmanned aerial vehicle to a ground testing device, dismounting an air propeller, switching on a direct-current power supply and starting the ground testing device;

B. the ground testing device starts automatic self-checking and calibration, and at the moment, a red light on the panel is on to remind a user of installing a coding disc for equal inertia replacement;

C. after the code disc is installed, an identity switch on the panel is toggled, the rotor enters the self-sustaining rotating speed after the indicator light turns green for 5s, and the rotor starts full-power acceleration in response to a step signal after 15 s. After the rotating speed is stable, the 'identification' switch is toggled again, the PWM signal value returns to zero, and the rotor starts to decelerate. After the rotor completely stops rotating, first-order linear link parameters can be obtained and displayed on a liquid crystal screen, and the indicator light turns red at the moment;

D. and (3) reinstalling the air propeller, turning on an identity switch, turning on the rotor to a self-sustaining rotating speed after an indicator light turns green for 5s, and automatically inputting PWM (pulse width modulation) signal values stabilized at 30%, 50% and 70% for the electric power system after 15s, 35s and 55s respectively. The "Identify" switch is toggled again, the PWM signal value returns to zero, and the rotor begins to decelerate. After the rotor completely stops rotating, the second power nonlinear term parameter can be obtained and displayed on the liquid crystal screen, and then the indicator light turns red;

E. when the work or the midway operation is abnormal, the Reset switch on the panel is turned, and the operation can return to the process before the process C.

F. If the power failure is interrupted halfway and the system is restarted or restarted, the system returns to the front of the flow B.

Claims (2)

1. The utility model provides a light-duty unmanned aerial vehicle electric power system identification device which characterized in that: the method comprises the following steps:

photoelectric coding disc: the ABS material disc is made of an ABS material, 6-9 notches are uniformly distributed on the circumference, and a compensation cylindrical block is designed near the center of the photoelectric coding disc;

a photoelectric sensor: the width of the photoelectric gate is 0.02m, when the optical path of the photoelectric gate is blocked, the signal pin of the photoelectric sensor outputs a high level, otherwise, the signal pin outputs a low level;

light-duty unmanned aerial vehicle electric power system ground test platform: the system consists of an AVR type singlechip minimum system, a bending moment type strain sensor, an A/D conversion element, a photoelectric sensor, a current-voltage sensor, a non-contact temperature sensor and a linear power rod, can measure parameters of tension, reaction torque, current, voltage, rotor temperature and rotating speed, can acquire output response data, and can identify and obtain time domain input and output mathematical model parameters of an electrodynamic force system;

the device has the working procedures as follows:

the method comprises the following steps: simplifying blades of an air propeller into a rectangle, calculating the rotational inertia of the propeller, reversely designing a disc-shaped coding disc, manufacturing the coding disc by using a material increase technology printer, setting the printing resolution to be maximum, and taking ABS as a material;

step two: installing a ground test board and a light unmanned aerial vehicle electric power system, using a photoelectric coding disc to replace a propeller, starting a motor by the light unmanned aerial vehicle electric power system, and stabilizing the rotating speed of the coding disc at the lowest self-sustaining rotating speed;

step three: inputting a step PWM signal value for an electrodynamic force system, wherein the upper limit of the step is a signal value corresponding to the maximum power, simultaneously sensing and recording a rotating speed increasing process by using a photoelectric sensor, determining a rotating speed increment according to time domain rotating speed response, finding a time increment corresponding to 63.2% of the rotating speed increment on a curve, wherein the time increment is a time constant of a first-order inertia link, and the ratio of the rotating speed increment to the signal value increment is a proportionality constant of the first-order inertia link;

step four: the coding disc is disassembled and the air propeller is reassembled;

step five: inputting a plurality of stable PWM signal values for a light unmanned aerial vehicle electric power system, sensing and recording the rotating speed by using a photoelectric sensor to obtain a plurality of stable rotating speed outputs, and fitting and calculating a second power parameter value by using a plurality of groups of input and output data;

step six: and (5) obtaining a time domain input and output mathematical model of the electric power system of the light unmanned aerial vehicle through sorting, and finishing the identification process.

2. An identification method of a light unmanned aerial vehicle electric power system identification device according to claim 1, wherein the identification method comprises the following steps: the method comprises the following steps:

step 1, building a nonlinear differential equation frame: a time domain mathematical model of a light unmanned aerial vehicle electrodynamic force system belongs to one of Hammerstein nonlinear differential equations and is composed of a first-order linear differential link and a second-order nonlinear link;

step 2, identifying linear link parameters: the photoelectric coding discs with equal rotational inertia replace an air propeller, so that the function of nondestructively simulating the inertia force of accelerated rotation is achieved, and the time constant and the amplification coefficient of a first-order linear differential link can be obtained by giving a step signal and responding according to the output rotating speed;

step 3, identifying nonlinear link parameters: the nonlinear link does not relate to differential terms, and parameters of the nonlinear link can be fit and calculated through multiple groups of steady-state input and output relations.

Images