CN105891698A - Boost circuit multi-parameter identification method - Google Patents

Abstract

The invention discloses a Boost circuit multi-parameter identification method. A Boost circuit comprises a constant-voltage source E, an inductor, an inductor equivalent series resistor, a filter capacitor, a filter capacitor equivalent series resistor, a power switch device S1, a diode S2, and a load resistor. The method comprises steps of: separately establishing the state equations of the S1 and the state equations of the S2 in different states according to the Boost circuit structure; combining the state equations in different states into a state space equation set to obtain a unified Boost circuit model; setting a sampling period, discretizing the state space equation set, and defining an observation matrix and a parameter matrix; sampling inductor current, output voltage, and the switch signal of a switching tube; acquiring the least square estimated value of the parameter matrix and computing an element parameter value required to be identified by using the parameter matrix acquired from the estimated value. The method may accurately identify the Boost circuit parameter in an ideal state.

Description

Boost circuit multi-parameter online identification method

Technical Field

The invention relates to a method for identifying circuit fault characteristic parameters, in particular to a method for identifying multiple parameters of a Boost circuit on line.

Background

System fault Prediction and Health Management (PHM) is a comprehensive fault detection, isolation and prediction and Health Management technique. The self health condition of the system is estimated by monitoring the fault characteristic parameters of the system and by means of various reasoning algorithms, the fault of the system can be monitored as early as possible before the fault of the system occurs, the degraded or fault part can be effectively predicted, the degraded or fault part can be accurately positioned, and a maintenance plan is given by combining various information resources, so that the situation-based maintenance and the autonomous guarantee of the system are realized, and the system has very important significance for reducing the maintenance cost, guaranteeing the reliability and the safety of the system, and improving the readiness and success rate of tasks. The PHM mainly comprises two parts of fault prediction and health management, wherein the fault prediction is the basis for realizing the system health management.

The application of power electronic technology can greatly improve the power density of the electric energy conversion device and reduce the volume and the weight. With the development of multi-electric and all-electric airplanes, the electricity consumption of airplanes is continuously increased, and the number of airborne power electronic devices is increased, so that higher requirements on the reliability, maintainability and testability of an airborne power electronic conversion device are provided, and the importance of the PHM of a power electronic system is increased.

Faults of power electronic conversion circuits can be mainly classified into structural faults and parametric faults according to different fault properties. The structural fault refers to a fault that a circuit topology changes due to short circuit or open circuit of a circuit device. Parametric faults refer to soft faults due to degradation of device parameters of the power electronic system. The parametric fault usually does not immediately lead to the system shutdown, but can cause the change of the output characteristic, so that the working performance and the reliability of the system are reduced; if the parametric fault can be predicted in time, the system can be prevented from developing into a worse system structural fault and from being seriously influenced by the structural fault, and the system reliability is greatly improved. Therefore, the key to realize the fault prediction is the accurate extraction of the characteristic parameters.

After the concept of discrete event dynamic system was proposed in the 80 s of the 20 th century, the hybrid system theory became a research hotspot in the field of control theory in recent years through years of research. A hybrid system refers to a system with continuous dynamic behavior and discrete event-driven behavior and the interaction of these two behaviors. Power electronic circuits, as switching power converters, are typically a hybrid dynamic system. The Matthew Senesky is based on the theory of a hybrid automaton, and provides a modeling method of a power electronic circuit based on a hybrid system.

The method comprises the steps that on the basis of the research of hybrid modeling in Zhejiang university, a hybrid system model of the power electronic circuit is constructed by utilizing a switching signal, inductive current and output voltage, and on the basis, the method for identifying the parameters of the power electronic circuit is obtained through a least square algorithm. However, the Boost circuit model thereof can cause the problems that the filter capacitor, the equivalent resistance thereof and the output resistance are difficult to identify in practical application.

Disclosure of Invention

The method takes the Boost circuit as an object, modifies the existing Boost hybrid model, and realizes the online identification of the multiple parameters of the Boost circuit, thereby providing a research basis for the fault prediction of the power electronic circuit.

The invention adopts the following technical scheme for solving the technical problems:

a Boost circuit multi-parameter online identification method is characterized by comprising the following steps:

the method comprises the steps of firstly, determining state variables of a Boost circuit, establishing state equations of the determined state variables in different switch tubular states according to the structure of the Boost circuit, and combining the established state equations into a Boost circuit space state equation set so as to obtain a Boost circuit model;

secondly, discretizing the Boost circuit model obtained in the first step, and then defining an observation matrix phi (t) and a parameter matrix theta_nWherein n is 1, 2;

thirdly, sampling the determined state variable and a switching signal of a switching tube to form an observation matrix phi (t);

and fourthly, obtaining a least square estimation value of the parameter matrix through a recursion algorithm, and calculating the parameter value of the element to be identified by using the parameter matrix obtained by the estimation value.

Further, the Boost circuit may equivalently include a constant voltage source E, an inductor and inductor equivalent series resistor, a filter capacitor and filter capacitor equivalent series resistor, and a power switch device S₁Diode S₂And a load resistor, wherein the negative electrode of the constant voltage source E is grounded, the positive electrode of the constant voltage source E is connected with one end of an inductor, the other end of the inductor is connected with one end of an inductor equivalent series resistor, and the other end of the inductor equivalent series resistor is simultaneously connected with the power switch device S₁And diode S₂Anode of (2), power switching device S₁Source electrode of (2) is grounded, two polePipe S₂The cathode of the filter capacitor is connected with the equivalent series resistor of the filter capacitor and the load resistor at the same time, the other end of the equivalent series resistor of the filter capacitor is connected with the filter capacitor, the other end of the filter capacitor is grounded, and the other end of the load resistor is grounded.

Further, in a first step, the output voltage u across the load resistor is adjusted_oAnd an inductor current i flowing through the inductor_LfDetermining the state variable as the state variable of the Boost circuit, and establishing the state variable u according to the structure of the Boost circuit_oAnd i_LfAt the switch tube S₁And a diode S₂Equation of state in different states:

state 1: switch tube S₁Conducting, diode S₂Turn off, note: s₁＝1，s₂When the voltage and current in the state are 0, the system of differential equations of the voltage and the current in the state is as follows:

$\left\{\begin{array}{c}{L}_f\·\frac{{\mathrm{di}}_{Lf}}{dt}=E-i_{Lf}{R}_L\\ C_f\frac{{\mathrm{du}}_{cf}}{dt}=C_f\frac{d(u_o-R_C\·(-u_o/R_o\left)\right)}{dt}=-u_o/R_o\end{array}\right.$

state 2: switch tube S₁Turn-off, diode S₂Open, record as: s₁＝0，s₂The differential equation of the voltage and current in this state is 1:

$\left\{\begin{array}{c}{L}_f\·\frac{{\mathrm{di}}_{Lf}}{dt}=E-u_o-i_{Lf}{R}_L\\ C_f\frac{{\mathrm{du}}_{cf}}{dt}=C_f\frac{d(u_o-R_C\·(i_{Lf}-u_o/R_o\left)\right)}{dt}=i_{Lf}-u_o/R_o\end{array}\right.$

state 3: switch tube S₁Turn-off, diode S₂Turn off, note: s₁＝0，s₂When the voltage and current in the state are 0, the system of differential equations of the voltage and the current in the state is as follows:

$\left\{\begin{array}{c}{L}_f\·\frac{{\mathrm{di}}_{Lf}}{dt}=0\\ C_f\frac{{\mathrm{du}}_{cf}}{dt}=C_f\frac{d(u_o-R_C\·(-u_o/R_o\left)\right)}{dt}=-u_o/R_o\end{array}\right.$

using switching signals s₁、s₂Writing the obtained equations under different switch tubular states into a uniform form:

$\left\{\begin{array}{c}{L}_f\·\frac{{\mathrm{di}}_{Lf}}{dt}=(s_1+s_2)(E-i_{Lf}{R}_L)-s_2{u}_o\\ C_f\frac{{\mathrm{du}}_{cf}}{dt}=C_f\frac{d(u_o-R_C\·(s_2{i}_{Lf}-u_o/R_o\left)\right)}{dt}=s_2{i}_{Lf}-u_o/R_o\end{array}\right.$

combining the obtained state equations into a state equation set to obtain a Boost circuit model as follows:

$\begin{array}{c}\left[\begin{array}{c}{\stackrel{\·}{i}}_{Lf}\\ {\stackrel{\·}{u}}_o\end{array}\right]=\left[\begin{array}{cc}-\frac{R_L}{L_f}& 0\\ 0& -\frac{1}{(R_C+R_o)C_f}\end{array}\right]\left[\begin{array}{c}{i}_{Lf}\\ u_o\end{array}\right]+s_2\left[\begin{array}{cc}0& -\frac{1}{L_f}\\ \frac{R_o(L_f-C_f{R}_C{R}_L)}{(R_C+R_o)L_f{C}_f}& -\frac{R_C{R}_o}{(R_C+R_o)L_f}\end{array}\right]\left[\begin{array}{c}{i}_{Lf}\\ u_o\end{array}\right]+s_1\left[\begin{array}{c}\frac{E}{L_f}\\ 0\end{array}\right]\\ +s_2\left[\begin{array}{c}\frac{E}{L_f}\\ \frac{R_C{R}_{o}E}{\left(R_C+R_o\right)L_f}\end{array}\right]+{\mathrm{\Δs}}_2\left[\begin{array}{c}0\\ \frac{R_C{R}_o}{(R_C+R_o)}\end{array}\right]i_{Lf}\end{array}$

wherein u is_o、i_LfRespectively output voltage and inductive current;differential quantities of the output voltage and the inductive current are respectively; s₁、s₂Separately switching tube S₁Diode S₂Switch signal quantity of, switch tube S₁At the time of conduction s₁1, switching tube S₁At turn-off time s₁0, diode S₂At the time of conduction s₂1, diode S₂At turn-off time s₂＝0；L_f、C_f、R_L、R_C、R_oRespectively obtaining an inductance value, a filter capacitance value, an inductance equivalent series resistance value, a filter capacitance equivalent series resistance value and a load resistance value in the Boost circuit;

in the second step, a sampling period T is set, and the Boost circuit model obtained in the second step is subjected to discretization treatment to obtain:

$\begin{array}{c}\left[\begin{array}{c}{i}_{Lf}\left(t\right)\\ u_o\left(t\right)\end{array}\right]=\left[\begin{array}{cc}1-\frac{R_L}{L_f}& 0\\ 0& 1-\frac{T}{(R_C+R_o)C_f}\end{array}\right]\left[\begin{array}{c}{i}_{Lf}\left(t-1\right)\\ u_o\left(t-1\right)\end{array}\right]+s_2\left(t-1\right)\left[\begin{array}{cc}0& -\frac{T}{L_f}\\ \frac{R_{o}T(L_f-C_f{R}_C{R}_L)}{(R_C+R_o)L_f{C}_f}+\frac{R_C{R}_o}{\left(R_C+R_o\right)}& -\frac{R_C{R}_{o}T}{(R_C+R_o)L_f}\end{array}\right]\left[\begin{array}{c}i\left(t-1\right)\\ u_o\left(t-1\right)\end{array}\right]+s_1\left(t-1\right)\left[\begin{array}{c}\frac{ET}{L_f}\\ 0\end{array}\right]\\ +s_2\left(t-1\right)\left[\begin{array}{c}\frac{ET}{L_f}\\ \frac{R_C{R}_{o}ET}{\left(R_C+R_o\right)L_f}\end{array}\right]-s_2\left(t-2\right)\left[\begin{array}{c}0\\ \frac{R_C{R}_o}{(R_C+R_o)}\end{array}\right]i\left(t-1\right)\end{array}$

defining an observation matrix as:

defining a parameter matrix as:

${\mathrm{\θ}}_1={\[1-\frac{R_L}{L_f},0,0,-\frac{T}{L_f},\frac{ET}{L_f},\frac{ET}{L_f},0\]}^T$

${\mathrm{\θ}}_2={\[0,1-\frac{T}{(R_C+R_o)C_f},\frac{R_{o}T(L_f-C_f{R}_C{R}_L)}{(R_C+R_o)L_f{C}_f}+\frac{R_C{R}_o}{(R_C+R_o)},-\frac{R_C{R}_{o}T}{(R_C+R_o)L_f},0,\frac{R_C{R}_{o}ET}{(R_C+R_o)L_f},\frac{R_C{R}_o}{(R_C+R_o)}\]}^T$

wherein u is_o、i_LfRespectively an output voltage value and an inductance current value; l is_f、C_f、R_L、R_C、R_oRespectively obtaining an inductance value, a filter capacitance value, an inductance equivalent series resistance value, a filter capacitance equivalent series resistance value and a load resistance value in the Boost circuit;

in the third step, the inductive current i at the t-1 moment is obtained according to the set sampling frequency_Lf(t-1), output voltage u_o(t-1) switching tube switching signal s₁(t-1), diode switch signal s₂(t-1) obtaining a switching signal s of the diode at the time of t-2₂(t-2) forming an observation matrix phi (t); obtaining the inductive current i at the moment t_Lf(t) output voltage u_o(t)；

In the fourth step, a parameter matrix theta is obtained through a recurrence algorithm_nLeast squares estimate of (d):

wherein n is 1, 2, x₁＝i_Lf(t)，x₂＝u_o(t)，θ_nThe estimated values of (c) are:

${\widehat{\mathrm{\θ}}}_1={\[a_{11},a_{12},a_{13},a_{14},a_{15},a_{16},a_{17}\]}^T$

${\widehat{\mathrm{\θ}}}_2={\[a_{21},a_{22},a_{23},a_{24},a_{25},a_{26},a_{27}\]}^T$

wherein,are each theta₁、θ₂An estimated value of (d); a is₁₁、a₁₂、a₁₃、a₁₄、a₁₅、a₁₆、a₁₇Is composed ofOf matrix coefficients, i.e. theta₁Each matrix coefficient estimate of a₂₁、a₂₂、a₂₃、a₂₄、a₂₅、a₂₆、a₂₇Is composed ofOf matrix coefficients, i.e. theta₂The estimated value of each matrix coefficient;

calculating the element parameter values to be identified according to the relationship between the parameter matrix obtained by estimation and the system parameters:

$\left\{\begin{array}{c}{L}_f=\frac{ET}{a_{15}}\\ R_L=\frac{L_f(1-a_{11})}{T}\\ R_o=\frac{a_{23}+a_{27}+a_{26}{R}_L/E}{1-a_{22}}\\ C_f=\frac{{\mathrm{TR}}_o-a_{26}{L}_f/E}{(a_{23}+a_{27}+a_{26}{R}_L/E)R_o}\\ R_C=\frac{a_{26}{L}_f}{(a_{23}+a_{27}+a_{26}{R}_L/E){\mathrm{EC}}_f}\end{array}\right.$

wherein L is_f、R_L、R_o、C_f、R_CThe inductance value, the equivalent series resistance value of the inductor, the load resistance value, the filter capacitance value and the equivalent series resistance value of the filter capacitor in the Boost circuit are respectively.

Compared with the prior art, the invention adopting the technical scheme has the following technical effects:

the online identification method provided by the invention has simple circuit, and improves the identification precision on the basis of not increasing electronic components;

after the scheme is adopted, the current detection point is subjected to inductive current and output voltage acquisition, the switching signals of the switching tube and the diode are switched on and off, and the filtering capacitance value C in the circuit can be realized on line by utilizing a recursion algorithm on the basis of a newly established Boost circuit model_fAnd its equivalent series resistance R_CFilter inductance value L_fAnd its equivalent series resistance R_LAnd a load resistance value R_oLeast squares estimation of (d);

the multi-parameter online identification method of the Boost circuit has the advantages that the switch switching of the diode is considered, and the filter capacitance C in the circuit is considered_fAnd its equivalent series resistance R_CResistance value R of load_oAnd (4) due to the identification influence, the established model is simultaneously suitable for modes of CCM and DCM.

Drawings

FIG. 1 is a schematic diagram of a Boost circuit structure in the method of the present invention;

FIG. 2 is a flow chart of the method of the present invention;

FIG. 3 is a circuit diagram of the present invention.

Detailed Description

The invention provides a Boost circuit multi-parameter online identification method, which aims to make the purpose, technical scheme and effect of the invention clearer and further describes the invention in detail by referring to the attached drawings and taking examples. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

A Boost circuit multi-parameter online identification method comprises the following steps:

first, determining the output voltage u_oAnd the inductor current i_LfAnd respectively establishing state equations of the state variables under different switch tubular states according to the Boost circuit structure for the state variables:

state 1: switch tube S₁Conducting, diode S₂Turn off, note: s₁＝1，s₂The differential of the voltage current in this state is given as 0The system of equations:

$\left\{\begin{array}{c}{L}_f\·\frac{{\mathrm{di}}_{Lf}}{dt}=E-i_{Lf}{R}_L\\ C_f\frac{{\mathrm{du}}_{cf}}{dt}=C_f\frac{d(u_o-R_C\·(-u_o/R_o\left)\right)}{dt}=-u_o/R_o\end{array}\right.$

state 2: switch tube S₁Turn-off, diode S₂Open, record as: s₁＝0，s₂The system of differential equations for voltage and current in this state is listed as 1:

$\left\{\begin{array}{c}{L}_f\·\frac{{\mathrm{di}}_{Lf}}{dt}=E-u_o-i_{Lf}{R}_L\\ C_f\frac{{\mathrm{du}}_{cf}}{dt}=C_f\frac{d(u_o-R_C\·(i_{Lf}-u_o/R_o\left)\right)}{dt}=i_{Lf}-u_o/R_o\end{array}\right.$

state 3: switch tube S₁Turn-off, diode S₂Turn off, note: s₁＝0，s₂The system of differential equations for voltage and current in this state is listed as 0:

$\left\{\begin{array}{c}{L}_f\·\frac{{\mathrm{di}}_{Lf}}{dt}=0\\ C_f\frac{{\mathrm{du}}_{cf}}{dt}=C_f\frac{d(u_o-R_C\·(-u_o/R_o\left)\right)}{dt}=-u_o/R_o\end{array}\right.$

second, using the switching signal s₁、s₂Writing the equations in different states obtained in the first step into a uniform form:

$\left\{\begin{array}{c}{L}_f\·\frac{{\mathrm{di}}_{Lf}}{dt}=(s_1+s_2)(E-i_{Lf}{R}_L)-s_2{u}_o\\ C_f\frac{{\mathrm{du}}_{cf}}{dt}=C_f\frac{d(u_o-R_C\·(s_2{i}_{Lf}-u_o/R_o\left)\right)}{dt}=s_2{i}_{Lf}-u_o/R_o\end{array}\right.$

and further arranging the data into a matrix form to obtain a model of the Boost circuit:

$\begin{array}{c}\left[\begin{array}{c}{\stackrel{\·}{i}}_{Lf}\\ {\stackrel{\·}{u}}_o\end{array}\right]=\left[\begin{array}{cc}-\frac{R_L}{L_f}& 0\\ 0& -\frac{1}{(R_C+R_o)C_f}\end{array}\right]\left[\begin{array}{c}{i}_{Lf}\\ u_o\end{array}\right]+s_2\left[\begin{array}{cc}0& -\frac{1}{L_f}\\ \frac{R_o(L_f-C_f{R}_C{R}_L)}{(R_C+R_o)L_f{C}_f}& -\frac{R_C{R}_o}{(R_C+R_o)L_f}\end{array}\right]\left[\begin{array}{c}{i}_{Lf}\\ u_o\end{array}\right]+s_1\left[\begin{array}{c}\frac{E}{L_f}\\ 0\end{array}\right]\\ +s_2\left[\begin{array}{c}\frac{E}{L_f}\\ \frac{R_C{R}_{o}E}{\left(R_C+R_o\right)L_f}\end{array}\right]+{\mathrm{\Δs}}_2\left[\begin{array}{c}0\\ \frac{R_C{R}_o}{(R_C+R_o)}\end{array}\right]i_{Lf}\end{array}$

and the third step specifically comprises the steps of setting the sampling period as T, and carrying out discretization treatment on the Boost circuit model obtained in the second step:

$\begin{array}{c}\left[\begin{array}{c}{i}_{Lf}\left(t\right)\\ u_o\left(t\right)\end{array}\right]=\left[\begin{array}{cc}1-\frac{R_L}{L_f}& 0\\ 0& 1-\frac{T}{(R_C+R_o)C_f}\end{array}\right]\left[\begin{array}{c}{i}_{Lf}\left(t-1\right)\\ u_o\left(t-1\right)\end{array}\right]+s_2\left(t-1\right)\left[\begin{array}{cc}0& -\frac{T}{L_f}\\ \frac{R_{o}T(L_f-C_f{R}_C{R}_L)}{(R_C+R_o)L_f{C}_f}+\frac{R_C{R}_o}{\left(R_C+R_o\right)}& -\frac{R_C{R}_{o}T}{(R_C+R_o)L_f}\end{array}\right]\left[\begin{array}{c}i\left(t-1\right)\\ u_o\left(t-1\right)\end{array}\right]+s_1\left(t-1\right)\left[\begin{array}{c}\frac{ET}{L_f}\\ 0\end{array}\right]\\ +s_2\left(t-1\right)\left[\begin{array}{c}\frac{ET}{L_f}\\ \frac{R_C{R}_{o}ET}{\left(R_C+R_o\right)L_f}\end{array}\right]-s_2\left(t-2\right)\left[\begin{array}{c}0\\ \frac{R_C{R}_o}{(R_C+R_o)}\end{array}\right]i\left(t-1\right)\end{array}$

defining an observation matrix:

defining a parameter matrix:

${\mathrm{\θ}}_1={\[1-\frac{R_L}{L_f},0,0,-\frac{T}{L_f},\frac{ET}{L_f},\frac{ET}{L_f},0\]}^T$

${\mathrm{\θ}}_2={\[0,1-\frac{T}{(R_C+R_o)C_f},\frac{R_{o}T(L_f-C_f{R}_C{R}_L)}{(R_C+R_o)L_f{C}_f}+\frac{R_C{R}_o}{(R_C+R_o)},-\frac{R_C{R}_{o}T}{(R_C+R_o)L_f},0,\frac{R_C{R}_{o}ET}{(R_C+R_o)L_f},\frac{R_C{R}_o}{(R_C+R_o)}\]}^T$

fourthly, obtaining the inductive current i at the t-1 moment according to the set sampling frequency_Lf(t-1), output voltage u_o(t-1) switching tube switching signal s₁(t-1), diode switch signal s₂(t-1) obtaining a switching signal s of the diode at the time of t-2₂(t-2) forming an observation matrix phi (t); obtaining the inductive current i at the moment t_Lf(t) output voltage u_o(t)。

Fifthly, obtaining a parameter matrix theta through a recursion algorithm_nLeast squares estimate of (d):

wherein n is 1 or 2; x is the number of₁＝i_Lf(t)，x₂＝u_o(t)；θ_nThe estimated values of (c) are:

${\widehat{\mathrm{\θ}}}_1={\[a_{11},a_{12},a_{13},a_{14},a_{15},a_{16},a_{17}\]}^T$ ${\widehat{\mathrm{\θ}}}_2={\[a_{21},a_{22},a_{23},a_{24},a_{25},a_{26},a_{27}\]}^T$

calculating the element parameter values to be identified according to the relationship between the parameter matrix obtained by estimation and the system parameters:

$\left\{\begin{array}{c}{L}_f=\frac{ET}{a_{15}}\\ R_L=\frac{L_f(1-a_{11})}{T}\\ R_o=\frac{a_{23}+a_{27}+a_{26}{R}_L/E}{1-a_{22}}\\ C_f=\frac{{\mathrm{TR}}_o-a_{26}{L}_f/E}{(a_{23}+a_{27}+a_{26}{R}_L/E)R_o}\\ R_C=\frac{a_{26}{L}_f}{(a_{23}+a_{27}+a_{26}{R}_L/E){\mathrm{EC}}_f}\end{array}\right.$

the Boost circuit mentioned herein has a schematic structure as shown in fig. 1, and includes a constant voltage source E and an inductor L_fAnd its equivalent series resistance R_LFilter capacitor C_fAnd its equivalent series resistance R_CDiode S₂Power switch device S₁And a load resistance R_o(ii) a Wherein, the negative pole of the constant voltage source E is grounded, and the positive pole of the constant voltage source E is connected with the inductor L_fOne end; inductor L_fThe other end is connected with an equivalent series resistance R of an inductor_LOne end of (a); inductance equivalent series resistance R_LThe other end of the first and second switches is connected with a power switch device S₁And diode S₂The anode of (1); power switch device S₁The source of (2) is grounded; diode S₂The cathode of the filter capacitor is simultaneously connected with an equivalent series resistor R of the filter capacitor_CAnd a load resistor R_o(ii) a Equivalent series resistance R of filter capacitor_CThe other end is connected with a filter capacitor C_f(ii) a Filter capacitor C_fThe other end of the first and second electrodes is grounded; load resistance R_oThe other end is grounded; wherein, the power switch device S₁May be a MOSFET or an IGBT.

The simulation circuit diagram of the Boost circuit multi-parameter online identification method provided by the invention is shown in fig. 3, and the simulation conditions, namely the converter parameters, are shown in table 1, wherein the constant voltage source E is 24V, and the inductance L is_fInductance equivalent series resistance R of 0.24mH_L0.4 Ω, capacitance C_f120 muF, filter capacitance equivalent series resistance R_C0.1 Ω, load resistance R_o15 Ω, switching frequency f_s50kHz, sampling frequency f_c5MHz, duty cycle D is 0.5. Firstly, carrying out circuit operation simulation, leading the inductive current, the output voltage and the diode voltage signal into the work space of the MATLAB after the operation is stopped, and carrying out minimum two through a recursion algorithmThe result of the multiplication is shown in Table 2, inductance L_fEstimated value is 0.24019mH, and inductance equivalent series resistance R_LIs 0.3954 omega, capacitance C_fEstimated value is 120.05 mu F, and filter capacitor equivalent series resistance R_CEstimated value of 0.1009 omega, load resistance R_oThe estimated value is 14.9964 omega, and the errors of the five are respectively 0.079%, 1.15%, 0.42%, 0.9% and 0.024%, so that the Boost circuit multi-parameter online identification method provided by the invention has high accuracy.

TABLE 1

E/V	Lf/mH	Cf/μF	Ro/Ω	RL/Ω	RC/Ω	fs/kHz	fc/MHz	D
									24	0.24	120	15	0.4	0.1	50	5	0.5

TABLE 2

	Lf/mH	RL/Ω	Cf/μF	RC/Ω	Ro/Ω
						Actual value	0.24	0.4	120	0.1	15
Estimated value	0.24019	0.3954	120.05	0.1009	14.9964
						Error/%)	0.079	1.15	0.042	0.9	0.024

The multi-parameter online identification method of the Boost circuit has the advantages that the switch switching of the diode is considered, and the filter capacitance C in the circuit is considered_fAnd its equivalent series resistance R_CResistance value R of load_oAnd (4) due to the identification influence, the established model is simultaneously suitable for modes of CCM and DCM. The model established in the method is more accurate and has higher identification precision than the model proposed by Zhejiang university.

It should be understood that equivalents and modifications of the technical solution and inventive concept thereof may occur to those skilled in the art, and all such modifications and alterations should fall within the scope of the appended claims.

Claims (3)

1. A Boost circuit multi-parameter online identification method is characterized by comprising the following steps:

A. determining state variables of a Boost circuit, establishing state equations of the determined state variables in different switch tubular states according to the structure of the Boost circuit, and combining the established state equations into a Boost circuit space state equation set so as to obtain a Boost circuit model;

B. discretizing the Boost circuit model obtained in the first step, and then defining an observation matrixAnd a parameter matrix theta_nWherein n is 1, 2;

C. sampling the determined state variables and the switching signals of the switching tubes to form an observation matrix

D. And obtaining a least square estimation value of the parameter matrix through a recursion algorithm, and calculating the element parameter value to be identified by using the parameter matrix obtained by the estimation value.

2. The method of claim 1, wherein the method comprises the steps of: the Boost circuit comprises a constant voltage source E, an inductor equivalent series resistor, a filter capacitor equivalent series resistor and a power switch device S₁Diode S₂And a load resistor, wherein the negative electrode of the constant voltage source E is grounded, the positive electrode of the constant voltage source E is connected with one end of an inductor, the other end of the inductor is connected with one end of an inductor equivalent series resistor, and the other end of the inductor equivalent series resistor is simultaneously connected with the power switch device S₁And diode S₂Anode of (2), power switching device S₁Is grounded, diode S₂The cathode of the filter capacitor is connected with the equivalent series resistor of the filter capacitor and the load resistor at the same time, the other end of the equivalent series resistor of the filter capacitor is connected with the filter capacitor, the other end of the filter capacitor is grounded, and the other end of the load resistor is grounded.

3. The method of claim 2, wherein the method comprises the steps of:

in step A, the output voltage u across the load resistor is measured_oAnd an inductor current i flowing through the inductor_LfDetermining the state variable as the state variable of the Boost circuit, and establishing the state variable u according to the structure of the Boost circuit_oAnd i_LfAt the switch tube S₁And a diode S₂Equation of state under different states, and obtaining the equation of stateAnd combining the state equation set to obtain a Boost circuit model as follows:

$\begin{array}{c}\left[\begin{array}{c}{\stackrel{\·}{i}}_{Lf}\\ {\stackrel{\·}{u}}_o\end{array}\right]=\left[\begin{array}{cc}-\frac{R_L}{L_f}& 0\\ 0& -\frac{T}{\left(R_C+R_o\right)C_f}\end{array}\right]\left[\begin{array}{c}{i}_{Lf}\\ u_o\end{array}\right]+s_2\left[\begin{array}{cc}0& -\frac{1}{L_f}\\ \frac{R_o\left(L_f-C_f{R}_C{R}_L\right)}{\left(R_C+R_o\right)L_f{C}_f}& -\frac{R_C{R}_o}{\left(R_C+R_o\right)L_f}\end{array}\right]\left[\begin{array}{c}{i}_{Lf}\\ u_o\end{array}\right]+s_1\left[\begin{array}{c}\frac{E}{L_f}\\ 0\end{array}\right]\\ +s_2\left[\begin{array}{c}\frac{E}{L_f}\\ \frac{R_C{R}_{o}E}{\left(R_C+R_o\right)L_f}\end{array}\right]+{\mathrm{\Δs}}_2\left[\begin{array}{c}0\\ \frac{R_C{R}_o}{\left(R_C+R_o\right)}\end{array}\right]i_{Lf}\end{array}$

wherein u is_o、i_LfRespectively output voltage and inductive current;differential quantities of the output voltage and the inductive current are respectively; s₁、s₂Separately switching tube S₁Diode S₂Switch signal quantity of, switch tube S₁At the time of conduction s₁1, switching tube S₁At turn-off time s₁0, diode S₂At the time of conduction s₂1, diode S₂At turn-off time s₂＝0；L_f、C_f、R_L、R_C、R_oRespectively obtaining an inductance value, a filter capacitance value, an inductance equivalent series resistance value, a filter capacitance equivalent series resistance value and a load resistance value in the Boost circuit;

in the step B, a sampling period T is set, and the Boost circuit model obtained in the second step is subjected to discretization treatment to obtain:

$\begin{array}{c}\left[\begin{array}{c}{i}_{Lf}\left(t\right)\\ u_o\left(t\right)\end{array}\right]=\left[\begin{array}{cc}1-\frac{R_L}{L_f}& 0\\ 0& 1-\frac{T}{\left(R_C+R_o\right)C_f}\end{array}\right]\left[\begin{array}{c}{i}_{Lf}\left(t-1\right)\\ u_o\left(t-1\right)\end{array}\right]+s_2\left(t-1\right)\left[\begin{array}{cc}0& -\frac{T}{L_f}\\ \frac{R_{o}T\left(L_f-C_f{R}_C{R}_L\right)}{\left(R_C+R_o\right)L_f{C}_f}+\frac{R_C{R}_o}{\left(R_C+R_o\right)}& -\frac{R_C{R}_{o}T}{\left(R_C+R_o\right)L_f}\end{array}\right]\left[\begin{array}{c}i\left(t-1\right)\\ u_o\left(t-1\right)\end{array}\right]+s_1\left(t-1\right)\left[\begin{array}{c}\frac{ET}{L_f}\\ 0\end{array}\right]\\ +s_2\left(t-1\right)\left[\begin{array}{c}\frac{ET}{L_f}\\ \frac{R_C{R}_{o}ET}{\left(R_C+R_o\right)L_f}\end{array}\right]-s_2\left(t-2\right)\left[\begin{array}{c}0\\ \frac{R_C{R}_o}{\left(R_C+R_o\right)}\end{array}\right]i\left(t-1\right)\end{array}$

defining an observation matrix as:

defining a parameter matrix as:

${\mathrm{\θ}}_1={\[1-\frac{R_L}{L_f},0,0,-\frac{T}{L_f},\frac{ET}{L_f},\frac{ET}{L_f},0\]}^T$

${\mathrm{\θ}}_2={\[0,1-\frac{T}{(R_C+R_o)C_f},\frac{R_{o}T(L_f-C_f{R}_C{R}_L)}{(R_C+R_o)L_f{C}_f}+\frac{R_C{R}_o}{(R_C+R_o)},-\frac{R_C{R}_{o}T}{(R_C+R_o)L_f},0,\frac{R_C{R}_{o}ET}{(R_C+R_o)L_f},\frac{R_C{R}_o}{(R_C+R_o)}\]}^T$

wherein u is_o、i_LfRespectively an output voltage value and an inductance current value; l is_f、C_f、R_L、R_C、R_oRespectively obtaining an inductance value, a filter capacitance value, an inductance equivalent series resistance value, a filter capacitance equivalent series resistance value and a load resistance value in the Boost circuit;

in step C, acquiring the inductive current i at the t-1 moment according to the set sampling frequency_Lf(t-1), output voltage u_o(t-1) switching tube switching signal s₁(t-1), diode switch signal s₂(t-1) obtaining a switching signal s of the diode at the time of t-2₂(t-2) forming an observation matrixObtaining the inductive current i at the moment t_Lf(t) output voltage u_o(t)；

In step D, a parameter matrix theta is obtained through a recursion algorithm_nLeast squares estimate of (d):

wherein n is 1, 2, x₁＝i_Lf(t)，x₂＝u_o(t)，θ_nThe estimated values of (c) are:

$\widehat{{\mathrm{\θ}}_1}={\[a_{11},a_{12},a_{13},a_{14},a_{15},a_{16},a_{17}\]}^T$

$\widehat{{\mathrm{\θ}}_2}={\[a_{21},a_{22},a_{23},a_{24},a_{25},a_{26},a_{27}\]}^T$

wherein,are each theta₁、θ₂An estimated value of (d); a is₁₁、a₁₂、a₁₃、a₁₄、a₁₅、a₁₆、a₁₇Is composed ofOf matrix coefficients, i.e. theta₁Each matrix coefficient estimate of a₂₁、a₂₂、a₂₃、a₂₄、a₂₅、a₂₆、a₂₇Is composed ofOf matrix coefficients, i.e. theta₂The estimated value of each matrix coefficient;

calculating the element parameter values to be identified according to the relationship between the parameter matrix obtained by estimation and the system parameters:

$\left\{\begin{array}{c}{L}_f=\frac{ET}{a_{15}}\\ R_L=\frac{L_f(1-a_{11})}{T}\\ R_o=\frac{a_{23}+a_{27}+a_{26}{R}_L/E}{1-a_{22}}\\ C_f=\frac{{\mathrm{TR}}_o-a_{26}{L}_f/E}{(a_{23}+a_{27}+a_{26}{R}_L/E)R_o}\\ R_C=\frac{a_{26}{L}_f}{(a_{23}+a_{27}+a_{26}{R}_L/E){\mathrm{EC}}_f}\end{array}\right.$

wherein L is_f、R_L、R_o、C_f、R_CThe inductance value, the equivalent series resistance value of the inductor, the load resistance value, the filter capacitance value and the equivalent series resistance value of the filter capacitor in the Boost circuit are respectively.