CN105260520A - Determination method for control algorithm in DC-DC converter modeling - Google Patents

Abstract

The invention relates to a control algorithm, in particular to a determination method for a control algorithm in DC-DC converter modeling. The determination method comprises a determination method for an incremental PID formula and a determination method for a PID control parameter. The determination method for the control algorithm in the DC-DC converter modeling, provided by the invention, is high in robustness, simple, feasible, easy to realize, and capable of enabling a control system for the DC-DC converter modeling to be high in stability and precision.

Description

The defining method of the control algolithm in DC-DC converter modeling

Technical field

The present invention relates to a kind of control algolithm, particularly relate to the defining method of the control algolithm in a kind of DC-DC converter modeling.

Background technology

The energy is exactly the material (the such as nuclear energy energy, photovoltaic energy, wind energy, the water conservancy energy, the mineral matter energy etc.) that can provide Conversion of Energy to nature.The energy is the material base of mankind's activity.Under current energy starved situation, the development of human society be unable to do without the appearance of high-grade energy and the use of advanced energy technology.The exploitation of the energy and the adjoint environmental problem of the energy, being the whole world, whole mankind's questions of common interest, is also the major issue of Chinese society economic development.Many countries also take a lot of active and effective measure for this reason and solve energy problem.But along with the consumption of the mineral matter energy such as coal prevailing in world's energy structure, oil, rock gas constantly increases, human society is faced with the test of more and more severeer energy and environment problem.Meanwhile, photovoltaic energy, as a kind of clean regenerative resource, starts the extensive concern causing people.Photovoltaic clean energy resource is that the luminous energy of the sun is converted into other forms of electric energy, heat energy or chemical energy, can not produce other harmful gas or solid wastes in energy sources conversion, is a kind of environmental protection, safe, clean novel energy.

World particularly developed country is attached great importance to for photovoltaic power generation technology, develops planning for it, increases input and is greatly developed.Volt generating utilizes this semiconductor devices of photovoltaic cell to absorb solar radiation energy, makes it the direct generation of electricity form changing into electric energy.As compared to the green generation technology of conventional power generation usage and other, photovoltaic generation has following advantage:

(1) sun power is inexhaustible, is irradiated to large 6000 times and generate electricity safe and reliable of energy that tellurian sun power consumes than the mankind, can not suffers the impact of energy crisis or fuel market instability;

(2) solar energy resources is available anywhere, and can power nearby.Need not long distance delivery, avoid the energy loss that long distance transmission line causes, also save Transmission Cost simultaneously.This is simultaneously also for domestic solar electricity generation system uses in western part of transmission of electricity inconvenience the condition of providing on a large scale;

(3) conversion process of energy of solar energy power generating is simple, is the conversion directly from photon to electronics, does not have pilot process (if thermal power transfer is mechanical energy, mechanical energy is converted to electromagnetic energy etc.) and mechanical motion, there is not mechanical wear.According to thermodynamic analysis, photovoltaic generation has very high theoretical generating efficiency, and can reach more than 80%, technical development potential is huge;

(4) photovoltaic generation does not produce any discarded object, and do not pollute, the public hazards such as noise, have no adverse effects to environment, are desirable clean energy resourcies;

(5) solar module structure is simple, and volume is little, lightweight, is convenient to transport and installs.The photovoltaic generating system construction period is short, and with changeable according to power load capacity, convenient, flexible, very easily combination, dilatation.

Summary of the invention

The features such as the present invention devises the defining method of the control algolithm in a kind of photovoltaic generating system in DC-DC converter modeling, has strong robustness, easy realization.

To achieve these goals, the present invention adopts following technical scheme: the defining method of the control algolithm in DC-DC converter modeling, is characterized in that: comprise the defining method of incremental PID formula and the defining method of pid control parameter;

The defining method of institute's incremental PID formula comprises:

(1) PID is controlled formula:

$\begin{array}{c}u=K_{e}e\left(t\right)+K_j{\∫}_0^{t}e\left(t\right)dt+K_d\frac{de\left(t\right)}{dt}+u_o\\ =K_p\[e\left(t\right)+\frac{1}{T_j}{\∫}_0^{t}e\left(t\right)dt+K_d\frac{de\left(t\right)}{dt}+u_o\]\end{array}---\left(1\right)$

In formula, u: the output of controller;

U ₀: the initial value of u when deviation is zero;

E (t): regulator input function, the i.e. deviation of specified rate and output quantity;

K _p: proportional gain, its inverse is called proportional band, i.e. δ=1/Kp;

T _i: integration time constant;

T _d: derivative time constant;

In differential equation discretize, and rewrite difference equation, be approximated as follows:

${\∫}_0^{t}e\left(t\right)dt={\Σ}_{j=0}^k{T}_e\left(j\right)---\left(2\right)$

$\frac{de\left(t\right)}{dt}=\frac{e\left(k\right)-e(k-1)}{T}---\left(3\right)$

In formula, T is the sampling period;

K is sampling sequence number;

E (k-1), e (k) are respectively (k-1) and kth time control cycle gained deviation;

(2) bring formula (2) and formula (3) into formula (1) to obtain:

$u\left(k\right)=K_p\{e\left(k\right)+\frac{T}{T_i}{\Σ}_{j=0}^{k}e\left(j\right)+\frac{T_d}{T}\[e\left(k\right)-e(k-1)\]\}+u_o---\left(4\right)$

Then the controlled quentity controlled variable u (k-1) in (k-1) moment is:

$u(k-1)=K_p\{e(k-1)+\frac{T}{T_i}{\Σ}_{j=0}^{k-1}e\left(j\right)+\frac{T_d}{T}\[e(k-1)-e(k-2)\]\}+u_o---\left(5\right)$

(3) formula (4)-Shi (5) obtains the increment of the controlled quentity controlled variable in kth moment and is:

$\begin{array}{c}\mathrm{\Δ}u\left(k\right)\\ =K_p\left\{e\left(k\right)-e\left(k-1\right)+\frac{T}{T_i}e\left(k\right)+\frac{T_d}{T}\[e\left(k\right)-2e\left(k-1\right)+e\left(k-2\right)\]\right\}\\ =K_p\left\{e\left(k\right)-e\left(k-1\right)+K_{i}e\left(k\right)+K_d\[e\left(k\right)-2e\left(k-1\right)+e\left(k-2\right)\]\right\}\end{array}---\left(6\right)$

In formula, for integral coefficient;

$K_d=K_p\frac{T_d}{T}$ For differential coefficient;

Formula (6) is incremental PID formula;

The defining method of described pid control parameter comprises:

A. select the sampling period Tmin that enough short, wherein Tmin should be less than 1/10th of object pure delay time.

B. ratio of being elected as by digitial controller controls, and the value of the p of scaling up COEFFICIENT K gradually, until the step response of system reaches threshold oscillation state;

C. degree of control is determined:

D. the degree of control determined according to step c is tabled look-up and is tried to achieve the value of each controling parameters.

E. make PID controller run according to the controling parameters of trying to achieve, and observe control effects, if system stability is inadequate, suitably can strengthen degree of control, then repeat steps d and e until obtain satisfied control effects.

The invention provides the defining method of the control algolithm in a kind of DC-DC converter modeling, strong robustness, simple, easily realize, can make the good stability of the control system of DC-DC converter modeling, precision is high.

Accompanying drawing explanation

Fig. 1 is the main circuit topological structure of DC-DC converter;

Fig. 2 is transducer equivalent electrical circuit when working in forward Boost;

Fig. 3 is transducer equivalent electrical circuit when working in forward Buck;

Fig. 4 is the modeling analysis of switch converters;

Fig. 5 is PID control system.

Embodiment

Below in conjunction with embodiment, technical scheme of the present invention is described in detail.

The defining method of the control algolithm in DC-DC converter modeling, is characterized in that: comprise the defining method of incremental PID formula and the defining method of pid control parameter;

The defining method of institute's incremental PID formula comprises:

(1) PID is controlled formula:

$\begin{array}{c}u=K_{e}e\left(t\right)+K_j{\∫}_0^{t}e\left(t\right)dt+K_d\frac{de\left(t\right)}{dt}+u_o\\ =K_p\[e\left(t\right)+\frac{1}{T_j}{\∫}_0^{t}e\left(t\right)dt+K_d\frac{de\left(t\right)}{dt}+u_o\]\end{array}---\left(1\right)$

In formula, u: the output of controller;

U ₀: the initial value of u when deviation is zero;

E (t): regulator input function, the i.e. deviation of specified rate and output quantity;

K _p: proportional gain, its inverse is called proportional band, i.e. δ=1/Kp;

T _i: integration time constant;

T _d: derivative time constant;

In differential equation discretize, and rewrite difference equation, be approximated as follows:

${\∫}_0^{t}e\left(t\right)dt={\Σ}_{j=0}^k{T}_e\left(j\right)---\left(2\right)$

$\frac{de\left(t\right)}{dt}=\frac{e\left(k\right)-e(k-1)}{T}---\left(3\right)$

In formula, T is the sampling period;

K is sampling sequence number;

E (k-1), e (k) are respectively (k-1) and kth time control cycle gained deviation;

(2) bring formula (2) and formula (3) into formula (1) to obtain:

$u\left(k\right)=K_p\{e\left(k\right)+\frac{T}{T_i}{\Σ}_{j=0}^{k}e\left(j\right)+\frac{T_d}{T}\[e\left(k\right)-e(k-1)\]\}+u_o---\left(4\right)$

Then the controlled quentity controlled variable u (k-1) in (k-1) moment is:

$u(k-1)=K_p\{e(k-1)+\frac{T}{T_i}{\Σ}_{j=0}^{k-1}e\left(j\right)+\frac{T_d}{T}\[e(k-1)-e(k-2)\]\}+u_o---\left(5\right)$

(3) formula (4)-Shi (5) obtains the increment of the controlled quentity controlled variable in kth moment and is:

$\begin{array}{c}\mathrm{\Δ}u\left(k\right)\\ =K_p\left\{e\left(k\right)-e\left(k-1\right)+\frac{T}{T_i}e\left(k\right)+\frac{T_d}{T}\[e\left(k\right)-2e\left(k-1\right)+e\left(k-2\right)\]\right\}\\ =K_p\left\{e\left(k\right)-e\left(k-1\right)+K_{i}e\left(k\right)+K_d\[e\left(k\right)-2e\left(k-1\right)+e\left(k-2\right)\]\right\}\end{array}---\left(6\right)$

In formula, for integral coefficient;

$K_d=K_p\frac{T_d}{T}$ For differential coefficient;

Formula (6) is incremental PID formula;

The defining method of described pid control parameter comprises:

A. select the sampling period Tmin that enough short, wherein Tmin should be less than 1/10th of object pure delay time.

B. ratio of being elected as by digitial controller controls, and the value of the p of scaling up COEFFICIENT K gradually, until the step response of system reaches threshold oscillation state;

C. degree of control is determined:

D. the degree of control determined according to step c is tabled look-up and is tried to achieve the value of each controling parameters.

E. make PID controller run according to the controling parameters of trying to achieve, and observe control effects, if system stability is inadequate, suitably can strengthen degree of control, then repeat steps d and e until obtain satisfied control effects.

DC-DC conversion is that fixing DC voltage conversion is become variable DC voltage, also referred to as DC chopped-wave.The working method of chopper has two kinds, and one is pulse width modulation mode, and the control signal cycle is constant, and change dutycycle, two is frequency modulation (PFM)s, changes control signal frequency.

The main circuit topology of the two-tube change of current type Buck-BoostDC-DC transducer that the present invention studies as shown in Figure 1, this transducer can realize unidirectional boosting and decompression transformation, control mode is regulate the dutycycle of the drive singal of metal-oxide-semiconductor Q1 and Q2 to realize buck to regulate, method is simple, and range of control is wide, can meet the voltage-regulation demand of relative broad range.C1 is electric capacity of voltage regulation, and L1 is afterflow inductance, and L2, C2 strobe.

When transducer works in forward Boost, the driving circuit driving work that Q2 is driven by PWM, Q1 keeps conducting state.This section first for Boost mode of operation, the operation mode of analytic transformation device and commutation course.For ease of analyzing, suppose that transducer reaches stable state, and all metal-oxide-semiconductors, diode and inductance are ideal component.Under Boost pattern, because Q1 is in conducting state, so circuit topology can be equivalent to Fig. 2:

When Q2 is in on-state, power supply Ui starts to inductance L charging, and charging current substantially constant is I _l, the voltage at electric capacity C two ends is to output terminal load supplying simultaneously, because C value is very large, substantially can keep output voltage U _oconstant, if the time that Q2 is in on-state is t _on, then the energy this stage inductance L stored is U _ii _lt _on.If the time that Q is in off-state is t _off, the U when Q2 is in off-state _ijointly charge to electric capacity C with L, and provide energy to output terminal load, the energy of inductance L release is during this period (U _o-U _i) I _lt _off, when circuit is in stable state, the energy that in one-period T, inductance L stores is equal with the energy of release, namely

U _iI _lt _on＝(U _O-U _i)I _lt _off(7)

Abbreviation obtains

$U_O=\frac{t_{on}+t_{off}}{t_{0ff}}{U}_i=\frac{T}{t_{off}}{U}_i---\left(8\right)$

In above formula output voltage U _ohigher than input voltage U _i, reach boosting effect.

When transducer works in forward Buck, the drive circuit works that Q1 is driven by PWM, Q2 maintains off-state.The topology of circuit can isoboles 3:

During Q1 is in on-state, if load current is i ₁, can following formula be obtained:

$L\frac{{\mathrm{di}}_1}{dt}+{\mathrm{Ri}}_1+E_M=E---\left(9\right)$

If this stage electric current initial value is I ₁₀, τ=L/R, brings above formula into and obtains:

$i_1=I_{10}{e}^{-\frac{t}{\mathrm{\τ}}}+\frac{E-E_M}{R}(1-e^{-\frac{t}{\mathrm{\τ}}})---\left(10\right)$

When Q1 is in off-state, if load current is i ₂, can following formula be obtained:

$L\frac{{\mathrm{di}}_2}{dt}+{\mathrm{Ri}}_2+E_M=0---\left(11\right)$

If this stage electric current initial value is I ₂₀, bring above formula into and obtain: t _on

$i_2=I_{20}{e}^{-\frac{t-ton}{\mathrm{\τ}}}-\frac{E_M}{R}(1-e^{-\frac{t-ton}{\mathrm{\τ}}})---\left(12\right)$

Have when current continuity:

I ₁₀＝i ₂(t ₂)(13)

I ₂₀＝i ₁(t ₁)(14)

Namely initial current value when Q1 enters on-state equals the current value at the end of the Q1 off-state stage, and conversely, initial current value when Q1 enters off-state equals the current value of Q1 at the end of the on-state stage.Can be drawn by formula (4), (6), (7), (8):

$I_{10}=\left(\frac{e^{\frac{t_1}{\mathrm{\τ}}}-1}{e^{\frac{T}{\mathrm{\τ}}}-1}\right)\frac{E}{R}-\frac{E_M}{R}=(\frac{e^{\mathrm{\α}\mathrm{\ρ}}-1}{e^{\mathrm{\ρ}}-1}-m)\frac{E}{R}---\left(15\right)$

$I_{20}=\left(\frac{1-e^{\frac{t_1}{\mathrm{\τ}}}}{1-e^{\frac{T}{\mathrm{\τ}}}}\right)\frac{E}{R}-\frac{E_M}{R}=(\frac{1-e^{\mathrm{\α}\mathrm{\ρ}}}{1-e^{\mathrm{\ρ}}}-m)\frac{E}{R}---\left(16\right)$

In formula i ₁₀and I ₂₀minimum value and the maximal value of load instantaneous electric current respectively.Can obtain with Taylor series approximation:

$I_{10}=I_{20}\≈\frac{(\mathrm{\α}-m)E}{R}=I_O---\left(17\right)$

Above formula illustrate smoothing reactor L infinitely great time load current complete straight time load current mean value I ₀, now load current maximal value and minimum value all equal mean value.In addition from the angle of energy, ignore the loss in circuit in one-period, the energy that power supply provides is equal with the energy of load consumption, that is:

${\mathrm{EI}}_O{t}_{on}={\mathrm{RI}}_O^{2}T=E_M{I}_{O}T---\left(18\right)$

Due to I _o=U _o/ R, obtains:

$U_O=\frac{t_{on}}{t_{on}+t_{off}}{U}_i=\frac{t_{on}}{T}{U}_i---\left(19\right)$

In above formula output voltage U _olower than input voltage U _i, reach antihypertensive effect.

In order to study the dynamic property of a switch converters and design its closed-loop control system, first need to set up its mathematical model.And for general switch converters, it is the nonlinear and time-varying system of a high-order, directly can not applies classical control theory and carry out analysis and design.

Include switching device in switch converters, the switching network that it is formed is discontinuous and nonlinear, and therefore, a set of analytical approach of linear continuous system can not directly apply in switch converters.General, the modeling method of switch converters can be divided into two large classes: Digital Simulation Method and parsing modeling, each class methods includes again several concrete grammar, and its exhaustive division situation is as shown in Figure 4.Here several conventional modeling method is simply introduced to studied transducer.

In switch converters modeling, Digital Simulation Method refers to and utilizes mathematical algorithm to carry out numerical evaluation to switch converters, obtains a kind of method of some characteristic of transducer, and it comprises direct numerical simulation method and indirect Digital Simulation Method.And Analytic modeling method refers to that employing analytical expression is to describe the modeling method of transducer characteristic, wherein most widely used a kind of Analytic modeling method is State-space Averaging Principle, be characterized in that a transducer each working stage in one-period makes average treatment, it is made to be equivalent to average continuous operation mode in one-period, thus adopt a continuous print differential equation to describe, by solving the differential equation, the steady-state characteristic of transducer and dynamic small-signal behaviour just can be obtained.The method physical concept is clear, and linear circuit and control theory can be utilized to carry out stable state and small-signal analysis to transducer, is designed with certain directive significance to transducer, but when disturbing signal frequency is higher, the accuracy of emulation has larger decline.

The basic thought of State-space Averaging Principle is: be averaging variable, be separated disturbance, linearization.Its general step is: 1. row write state equation stage by stage; 2. quiescent point is asked; 3. ac small signal state equation and output equation is set up.

Control algolithm is a very important link for Control System Design, because only could control accurately output voltage electric current in system after determining algorithm, the quality of algorithm is directly connected to the height of Control system resolution.Native system is through and regulates the dutycycle of metal-oxide-semiconductor to control output voltage electric current, in the current and voltage feedback control procedure of DC-DC converter, proportionally (P), integration (I), differential (D) three controlling units are the control modes be most widely used at present to the PID Closed loop Control that it controls.It has the advantages such as strong robustness, easily realization.

Current control system mainly can be divided into analog control system and digital control system two kinds, compared with analog control system, digital control system have structure simple, more complicated control law can be realized, accurately control the advantages such as controlled device.In view of above factor, this design alternative digital control system is that core realizes PID control controlled device with arm processor.

Digital control system is primarily of digital control processor and closed-loop control system composition, the output of controlled device and control signal are generally continuous print simulating signal, and processor accepts and the signal of process and output is binary-coded digit signal, therefore needs to carry out digital-to-analogue A/D and modulus D/A to signal and changes.The control object of the design is the dutycycle of metal-oxide-semiconductor, and then reach the effect controlling output voltage electric current, therefore need to be converted to digital signal through A/D to process measured, then the digital signal processed is converted to simulating signal through D/A controls metal-oxide-semiconductor.

(1) PID control principle: in PID control procedure, does not generally need the mathematical model setting up out controlled device, only needs rule of thumb to adjust closed loop controller parameter and just can obtain comparatively satisfied control effects [27].Adopt the control system of PID controller as shown in Figure 5.As can be seen from the figure, ratio (P), integration (I), differential (D) three controlling units are relations in parallel, can regulate respectively three parameters, three kinds of control laws can use simultaneously, also only can adopt one or both control laws wherein.

PID controls the corresponding formula that controls:

$u=K_{e}e\left(t\right)+K_j{\∫}_0^{t}e\left(t\right)dt+K_d\frac{de\left(t\right)}{dt}+u_o$

$=K_p\[e\left(t\right)+\frac{1}{T_j}{\∫}_0^{t}e\left(t\right)dt+T_d\frac{de\left(t\right)}{dt}+u_o\]---\left(20\right)$

In formula, u: the output of controller;

U0: the initial value of u when deviation is zero;

E (t): regulator input function, the i.e. deviation of specified rate and output quantity;

Kp: proportional gain, its inverse is called proportional band, i.e. δ=1/Kp;

Ti: integration time constant;

Td: derivative time constant.

Ratio controls

It is one control law more intuitively that ratio controls, and the variable quantity of control action is directly proportional to inclined extent, that is:

Δu＝u-u _o＝K _ee(t)(21)

In formula, u is the output of controller; The initial value of u0 is deviation when being zero u; E (t) is deviation; Ke is proportional gain.Ke is larger, and the transient process of system is faster, can reach and eliminate static error effect rapidly; But Ke is excessive, easily make system overshoot, produce vibration, cause instability.Therefore, suitable selection percentage coefficient can make system reach settling time short while ensure again stable effect.

Integration control

Integration control, refers to that the variable quantity of control action and deviation are integrated into direct ratio to the time, namely

$\mathrm{\Δ}u=u-u_o=K_i{\∫}_0^{t}e\left(t\right)dt---\left(22\right)$

The effect of integration control is: as long as system exists error, and the effect of integration control will constantly accumulate, and constantly export controlled quentity controlled variable to eliminate error, as long as therefore have time enough, integration control just can eliminate error completely.But fruit is as too strong in the effect of integration control, and system overshoot can be caused excessive, and even make system occur vibration, therefore integration control is seldom used alone, and generally should control to use in conjunction with differential.

Differential controls

Differential controls to refer to that the variable quantity of control action is directly proportional to the pace of change of deviation, namely

$\mathrm{\Δ}u=u-u_o=K_d\frac{de\left(t\right)}{dt}---\left(23\right)$

The differential action has anticipation in phase place with on the time, depends on the variation tendency of deviation, and is carry out controlling according to the numerical value of deviation own.Differential controls the dynamic responding speed not only can accelerating system, reduces regulation time, effectively improves the dynamic property of system, can also reduce system overshoot simultaneously, overcome system oscillation, improves the stability of system.

(2) Digital PID Algorithm: realize pid control algorithm for the ease of computing machine, differential equation discretize, must be rewritten into difference equation, can do as lower aprons for this reason:

${\∫}_0^{t}e\left(t\right)dt={\Σ}_{j=0}^k{T}_e\left(j\right)---\left(24\right)$

$\frac{de\left(t\right)}{dt}=\frac{e\left(k\right)-e(k-1)}{T}---\left(25\right)$

In formula, T is the sampling period; K is sampling sequence number; E (k-1), e (k) are respectively (k-1) and kth time control cycle gained deviation.

According to above formula, difference equation can be obtained

$u\left(k\right)=K_p\{e\left(k\right)+\frac{T}{T_i}{\Σ}_{j=0}^{k}e\left(j\right)+\frac{T_d}{T}\[e\left(k\right)-e(k-1)\]\}+u_o---\left(26\right)$

In formula, Section 1 plays ratio control action, and Section 2 plays integral control action, and Section 3 plays derivative control action.U (k) is full dose output, and it corresponds to the position that each sampling instant of controlled device topworks should reach, and is therefore also called Position Form PID formula.

As can be seen from formula Position Form PID formula, it directly calculates and inconvenience in a computer, because will add up e (j), need to take more storage unit, calculated amount is excessive, and is not easy to calculate.

Be not difficult to write out the controlled quentity controlled variable u (k-1) in (k-1) moment according to formula position type PID formula, namely

$u(k-1)=K_p\{e(k-1)+\frac{T}{T_i}{\Σ}_{j=0}^{k-1}e\left(j\right)+\frac{T_d}{T}\[e(k-1)-e(k-2)\]\}+u_o----\left(27\right)$

The increment that two formulas subtract each other the controlled quentity controlled variable obtaining the kth moment is

$\begin{array}{c}\mathrm{\Δ}u\left(k\right)\\ =K_p\left\{e\left(k\right)-e\left(k-1\right)+\frac{T}{T_i}e\left(k\right)+\frac{T_d}{T}\[e\left(k\right)-2e\left(k-1\right)+e\left(k-2\right)\]\right\}\\ =K_p\left\{e\left(k\right)-e\left(k-1\right)+K_{i}e\left(k\right)+K_d\[e\left(k\right)-2e\left(k-1\right)+e\left(k-2\right)\]\right\}\end{array}---\left(28\right)$

In formula, for integral coefficient, for differential coefficient.

The result that above formula calculates reflect the kth time of controller and kth export for-1 time between increment, so be referred to as incremental PID formula.

Increment type controls to control algorithmically to have done a bit to improve relative to position model, makes computation amount.In addition, be increment due to what export after processor calculating, so the impact caused system when producing misoperation is very little, the method that utilogic judges if desired is removed; Owing to only exporting the increment that this controls, controlled quentity controlled variable and initial position have nothing to do, thus be conducive to realizing manually and automatic between switching undisturbed; And can not cumulative errors be caused in computation process.

In practical engineering application, control mode should be selected according to the actual conditions of controlled device.It is generally acknowledged, require, in higher system, Position Form PID algorithm should be adopted in control accuracy; And be in the system of performer with metal-oxide-semiconductor, should increment type PID algorithm be adopted.Because this design controls metal-oxide-semiconductor, accuracy requirement is not very high, so adopt incremental timestamp.

For improving the control accuracy of system, can improve on the basis of standard P ID control algolithm, improve the performance of digitial controller, mainly contain following several method: the incomplete differential method for differential term, the integration partition method for integration item, integration margining amplitude technique, differential forward and input filter and shift integral method etc.In conjunction with the feature of DC-DC converter, the design adopts the control method of integral separating PID to improve the performance of controller.

In pid number control system, introduce integral element, main object eliminates static error, raising precision.The output voltage of the design's system and electric current likely can cause numerical value to suddenly change because of the factor such as external interference and starting when system is normally run, system can produce very large instantaneous output bias, the integration of integration control in PID arithmetic is caused to accumulate, the controlled quentity controlled variable causing it to calculate exceedes the limit controlled quentity controlled variable corresponding to topworks's maximum actuation scope, finally cause system to produce larger overshoot, even cause the vibration of system.In PID arithmetic, introduce integration partition method, both maintained integral control action, turn reduce system overshoot simultaneously, make Systematical control obtain larger improvement.

The basic thought of integration partition method does not carry out integration when deviation is large, only just carries out integration accumulation when the absolute value of deviation is less than the threshold values ε of a reservation, and ε is the error of the difference of output voltage electric current and given target voltage electric current herein.That is:

| e (k) | < ε is also deviate | e (k) | time larger, adopt PID to control, system so both can have been avoided to produce excessive overshoot, system can be made again to have response speed faster.

| e (k) | > ε is also deviate | e (k) | time smaller, adopt PD to control, the control accuracy of system can be ensured.

Utilize the logical operation function of microprocessor, the process of integral separating PID control can be determined easily.After considering integration introjection, controller gain presents increase tendency on the whole, therefore will by corresponding for proportional gain reduction, so that keeping system stability substantially constant [28].

(3) the adjusting of PID controller parameter

Consider the factor such as algorithm speed and processor speed, the design adopts incremental timestamp algorithm in order to eliminate the system overshoot and oscillatory occurences that are caused by integration item, have employed integral separating PID control method.

The adjustment of Digital PID Controller parameter is adjusted, and needs test parameter being carried out to cambic calculating and repeatability, sometimes needs to rely on the commissioning experience of debugging person's accumulation to debug, and could obtain satisfied effect of adjusting.Therefore, the adjustment of Digital PID Controller parameter is adjusted is a very complicated job.

The method that PID controller parameter is adjusted is more, and current common method mainly contains following two kinds:

A () theoretical calculation: only have and obtain controlled device mathematical model accurately, can use theoretical calculation CONTROLLER DESIGN parameter, because process is complicated, calculated amount is large, and this point is difficult to accomplish in working control process.

(b) engineering turning method: applying maximum in engineering turning method is expanding critical proportion method.The method is applicable to the controlled device of self-equilibrium characteristic, step of specifically adjusting following [29]:

1. the sampling period Tmin that enough short is selected.General Tmin should be less than 1/10th of object pure delay time.

2. ratio of being elected as by digitial controller controls, and the value of the p of scaling up COEFFICIENT K gradually, until the step response of system reaches threshold oscillation state.

3. degree of control is determined:

Degree of control reflects the relative extent of numerically controlled effect relative to the effect of analogue enlargement.Because degree of control is a relative concept only representing relation control effect, so for two square-error areas, do not need to be calculated in practical engineering application.Generally when the value of degree of control is 1.05, just can think that the effect of numerically controlled effect and analogue enlargement is suitable.

4. the degree of control determined according to previous step is tabled look-up and is tried to achieve the value of each controling parameters.

5. make PID controller run according to the controling parameters of trying to achieve, and observe control effects, if system stability is inadequate, suitably can strengthen degree of control, then repeat 4. with 5. until obtain satisfied control effects.What current application was more is engineering turning method.Do not rely on the mathematical model of object when the great advantage of this method is exactly parameter tuning, without the need to obtaining the mathematical model of controlled device, directly can carry out scene in the controls and adjusting, simple.

Claims (1)

1. The defining method of the control algolithm in the modeling of 1.DC-DC transducer, is characterized in that: comprise the defining method of incremental PID formula and the defining method of pid control parameter;

   The defining method of institute's incremental PID formula comprises:

   (1) PID is controlled formula:

   In formula, u: the output of controller;

   U ₀: the initial value of u when deviation is zero;

   E (t): regulator input function, the i.e. deviation of specified rate and output quantity;

   K _p: proportional gain, its inverse is called proportional band, i.e. δ=1/Kp;

   T _i: integration time constant;

   T _d: derivative time constant;

   In differential equation discretize, and rewrite difference equation, be approximated as follows:

   In formula, T is the sampling period;

   K is sampling sequence number;

   E (k-1), e (k) are respectively (k-1) and kth time control cycle gained deviation;

   (2) bring formula (2) and formula (3) into formula (1) to obtain:

   Then the controlled quentity controlled variable u (k-1) in (k-1) moment is:

   (3) formula (4)-Shi (5) obtains the increment of the controlled quentity controlled variable in kth moment and is:

   In formula, for integral coefficient;

   for differential coefficient;

   Formula (6) is incremental PID formula;

   The defining method of described pid control parameter comprises:

   A. select the sampling period Tmin that enough short, wherein Tmin should be less than 1/10th of object pure delay time.

   B. ratio of being elected as by digitial controller controls, and the value of the p of scaling up COEFFICIENT K gradually, until the step response of system reaches threshold oscillation state;

   C. degree of control is determined:

   D. the degree of control determined according to step c is tabled look-up and is tried to achieve the value of each controling parameters.

   E. make PID controller run according to the controling parameters of trying to achieve, and observe control effects, if system stability is inadequate, suitably can strengthen degree of control, then repeat steps d and e until obtain satisfied control effects.