CN103970961A - Temperature sensor-free LED junction temperature predicting and controlling method - Google Patents

Abstract

The invention discloses a temperature sensor-free LED junction temperature predicting and controlling method. The method comprises the steps of S1, obtaining a simplified mathematical model and parameters of an LED according to a sample experimental test curve; S2, predicting the junction temperature, Tj prediction, of the LED by means of the simplified mathematical model; S3, predicting the output luminous flux, fai (Tj prediction, Ij), of the LED at the moment according to the predicted junction temperature obtained in the S2; S4, adjusting and updating the driving current If according to the predicted junction temperature and a preset condition of the output luminous flux; S5, repeating the predicting and controlling process. The LED junction temperature predicting method is reliable and effective. Furthermore, due to the adoption of a sensor-free control structure, the number of system elements used is reduced, system reliability is improved, and the method has great advantages in online testing.

Description

A kind of prediction and control method of the LED junction temperature without temperature sensor

Technical field

The present invention relates to a kind of field of illuminating lamps, be specifically related to a kind of prediction and control method of the LED junction temperature without temperature sensor.

Background technology

Light emitting diode (LED) is the semi-conductor electricity optical device that utilizes gallium nitride (GaN) material to prepare, and realizes the conversion of electric energy to luminous energy.As electric light source technology of new generation, LED illumination has higher efficiency with respect to traditional electrical light source (incandescent lamp and fluorescent light), according to statistics: electric consumption on lighting accounts for nearly 20% of total electricity consumption, if extensively adopt LED to replace conventional light source for indoor and outdoor lighting, can save every year nearly 40% illumination electric energy consumption.In addition the advantage that, LED also has life-span length, fast response time and pollutes without noxious materials such as mercurys.Therefore, LED illumination has become common recognition as most potential high-efficiency environment friendly energy-conservation technology.

But, LED electric light source is unlike traditional electrical light source, can be directly connected on the electric main of 220V, but need to design special low voltage drive circuit, and very high to the performance requirement of driving power, the performance of LED light source is extremely responsive to factors such as environment temperatures, shows as chromaticity coordinates fluctuation, colour temperature fluctuation and illumination fluctuation etc.In fact, the as easy as rolling off a log threshold value that exceedes human eye permission of fluctuation of these lighting engineering indexs, will produce 0.001 chromaticity coordinates and change as the junction temperature of LED only changes 6 DEG C, and the chromaticity coordinates threshold value of human eye only has an appointment 0.0005, thereby will cause that people produces dazzle, headache to LED light source, even physiological function disorder waits bad reaction.LED light source unstable greatly affected the quality of illumination, even affects the health of lighting object, becomes a difficult point of LED Lighting Industry high-end designs.

In recent years, the variations injunction temperature of LED becomes an effective means that quantizes LED performance gradually.LED junction temperature refers to the temperature of quantum hydrazine recombination region in LED chip, and the volume of LED quantum hydrazine recombination region is very little, is generally nanometer (nm) × square millimeter (mm ²) order of magnitude, the measuring method of LED junction temperature often adopts forward voltage method, equivalent thermal resistance method, responsive to temperature optical parametric method and infrared thermal imaging etc.But these methods are mainly used in laboratory condition, and can not be used in the site of deployment of LED lighting source.In application, conventionally utilize cheap temperature sensor to carry out detection and the control of LED light source junction temperature at the scene, and then realize the control of light output, the possibility that prevents too high junction temperature safeguard measure is provided simultaneously.But compared with using longevity with LED light source, this temperature sensor life-span is not long, may greatly affect the reliability of whole illuminator.

Summary of the invention

Given this, the present invention proposes a kind of prediction of the junction temperature without sensor LED and control method of novelty on the hardware foundation of the intelligent drives of LED.The method proposes on single-chip LED test feature and mathematics physics model basis, its theoretical foundation is that drive current-luminous flux characteristic of LED is set up to a kind of mathematics physics model of approximate simplification, realizes prediction and the control of the junction temperature to LED on the basis of this model.

The present invention realizes a kind of prediction and control method of the LED junction temperature without temperature sensor by following technical scheme, comprise the following steps: S1., by sample experiment test curve, obtains simplification mathematical model and the parameter thereof of LED; S2. utilize the junction temperature T that simplifies mathematical model prediction LED _{j prediction}; S3. according to the junction temperature of predicting in S2, predict the now output light flux φ (T of LED _{j prediction}, I _j); S4. according to the junction temperature of prediction and output light flux pre-conditioned, adjust and upgrade drive current I _f; S5. repeat the process of above prediction and control.

Further, consider environment temperature T _awith drive current I _funder condition, junction temperature T in described step S2 _{j prediction}to predict by equation (1):

Wherein, η _eQEfor external quantum efficiency, R _jafor the PN junction of LED chip is to the thermal resistance of environment, K _vcfor temperature coefficient, T _j0represent initial junction temperature, V _f(T _j0) expression junction temperature is T _j0time forward voltage, I _ffor drive current, T _afor environment temperature.

Further, junction temperature is T _j0time forward voltage V _f(T _j0) I relevant to drive current _f, adopt following methods correction, that is, adopt equation (2) to simplify forward voltage and forward current relationship,

V _f(T _j0)＝I _f·R _s+α ₀(2)

R _sfor equivalent string resistance, α ₀for translational movement, characteristic obtains by experiment,

Due to R _swith electric current I _fincrease and increase, adopt electric current I ₀equivalent series resistance R under condition _s(I ₀) modified R _s,

R _s(I _f)＝R _s(I ₀)-α ₁(I _f-I ₀) (3)

Correction factor α ₁test obtains by experiment.

Further, temperature coefficient K in described equation (1) _vcto obtain by equation (4):

$K_{\mathrm{vc}}=\frac{V_f\left(T_j\right)-V_f\left(T_{j0}\right)}{T_j-T_{j0}}---\left(4\right).$

Further, the middle junction temperature of described equation (1) is to the thermal resistance R of environment _jaobtain by equation (5):

R _ja(I _f)＝R _ja(I ₀)-α ₂(I _f-I ₀) (5)

R _ja(I ₀) record drive current I=I under laboratory condition ₀time LED thermal resistance, R _ja(I _f) be drive current I=I _fthe thermal resistance value of Shi Xiuzheng, correction factor α ₂by the R to LED in different forward current I situations _jawith environment temperature T _afamily curve carry out matching and obtain.

Further, external quantum efficiency η in described equation (1) _eQEobtain by equation (6):

${\mathrm{\η}}_{\mathrm{EQE}}={\mathrm{\α}}_3+{\mathrm{\α}}_4{I}_f+{\mathrm{\α}}_5{I}_f^2---\left(6\right)$

Factor alpha ₃, α ₄and α ₅by the I-V characteristic of LED with I-L characteristic test curve carries out curve fitting or peak fitting obtains.

Further, in described step S3, luminous flux prediction realizes by equation (7)

φ _L(T _j,I _j)＝φ _L(T _j0,I _j)·(α ₆(T _j-T _j0)+α ₇) (7)

I _jexpression junction temperature is T _jtime drive current, factor alpha ₆, α ₇can be obtained by forward electric current-luminous flux characteristic curve fitting.

Owing to having adopted technique scheme, the present invention has advantages of as follows:

The designed LED junction temperature Forecasting Methodology of the present invention is reliable and effectively, and due to the control structure adopting without sensor, has reduced the usage quantity of system element, has improved the reliability of system, has greatly online detection advantage.

Brief description of the drawings

In order to make the object, technical solutions and advantages of the present invention clearer, below in conjunction with accompanying drawing, the present invention is described in further detail, wherein:

Fig. 1 is the simplified model block scheme of electric current-luminous flux model;

Fig. 2 is the typical test curve figure of forward electric current-forward voltage;

Fig. 3 is temperature-forward voltage characteristic curve map;

Fig. 4 is in forward current I situation, LED R _jaand T ₀experimental features curve map;

Fig. 5 is electric current-external quantum efficiency graph of relation;

Fig. 6 is forward current-luminous flux characteristic test curve map;

Fig. 7 is the output light flux control chart based on temperature sensor;

Fig. 8 is the output light flux control chart without sensor;

Fig. 9 is the method for the invention process flow diagram.

Embodiment

Below with reference to accompanying drawing, the preferred embodiments of the present invention are described in detail; Should be appreciated that preferred embodiment is only for the present invention is described, instead of in order to limit the scope of the invention.

The junction temperature Forecasting Methodology that the present invention relates to a kind of LED without sensor, the method proposes on single-chip LED test feature and mathematics physics model basis, and its theoretical foundation is the drive current (I to LED _f)-luminous flux (L) characteristic is set up a kind of mathematics physics model of approximate simplification, realizes prediction and the control of the junction temperature to LED on the basis of this model.As shown in Figure 1, this model is with the drive current (I of LED for the Establishing process of model _f) be input variable, investigate the operating ambient temperature (T of LED simultaneously _a) impact, prediction LED junction temperature (T _j), and corresponding luminous flux output (L).

Under laboratory condition, forward electric current I=I _fdo the used time, the junction temperature T of LED _jthe normal forward voltage method that adopts is measured estimation, and detailed process is as follows.

First test temperature T 0 (such as environment temperature, the now initial junction temperature T of LED _j0just equal initial temperature T ₀, in calculating afterwards by T _j0use T ₀represent) and in short-term little electric current (1ms/100uA) condition under, the forward voltage V that LED is corresponding _f(T ₀), then the drive current I that reaches hot stable state is introduced to identical in short-term little electric current, and record the forward voltage V that LED is corresponding _f(T _j), the junction temperature T under drive current I condition _jcan be expressed as,

$T_j=\frac{V_f\left(T_j\right)-V_f\left(T_0\right)}{K_{\mathrm{vc}}}+T_0---\left(8\right)$

Further, under drive current I effect, the junction temperature T of LED _jalso available junction temperature is to the thermal resistance R of environment _jadefine, that is,

$R_{\mathrm{ja}}=\frac{T_j-T_a}{P_{\mathrm{th}}}---\left(9\right)$

Wherein, T _afor environment temperature, P _ththe heating power that represents LED, can be approximately equal to electric power (P _e=I _fv _f) poor with luminous power;

According to (9) formula, can obtain

$T_j=\frac{P_{\mathrm{th}}}{1/R_{\mathrm{ja}}}+T_a---\left(10\right)$

If the external quantum efficiency of LED is η _eQE,

P _th＝(1-η _EQE)·P _e＝(1-η _EQE)·I _f·V _f(11)

(11) formula is brought into (10) formula, can obtain

$T_j=\frac{P_{\mathrm{th}}}{1/R_{\mathrm{ja}}}+T_a=\frac{(1-{\mathrm{\η}}_{\mathrm{EQE}})\·I_f\·V_f}{1/R_{\mathrm{ja}}}+T_a---\left(12\right)$

(8) formula is brought into (12) formula, can obtain

$T_j=\frac{(1-{\mathrm{\η}}_{\mathrm{EQE}})\·I_f\·V_f}{1/R_{\mathrm{ja}}}+T_a=\frac{(1-{\mathrm{\η}}_{\mathrm{EQE}})\·I_f}{1/R_{\mathrm{ja}}}\·[k_{\mathrm{vc}}(T_j-T_0)+V_f\left(T_0\right)]+T_a\⇒$

$T_j[1-\frac{(1-{\mathrm{\η}}_{\mathrm{EQE}})\·I_f\·k_{\mathrm{vc}}}{1/R_{\mathrm{ja}}}]=T_a-\frac{(1-{\mathrm{\η}}_{\mathrm{EQE}})\·I_f\·k_{\mathrm{vc}}\·T_0}{1/R_{\mathrm{ja}}}+\frac{(1-{\mathrm{\η}}_{\mathrm{EQE}})\·I_f\·V_v\left(T_0\right)}{1/R_{\mathrm{ja}}}\⇒$

$T_j=\frac{T_a-\frac{(1-{\mathrm{\η}}_{\mathrm{EQE}})\·I_f}{1/R_{\mathrm{ja}}}\·K_{\mathrm{vc}}\·T_0+\frac{(1-{\mathrm{\η}}_{\mathrm{EQE}})\·I_f}{1/R_{\mathrm{ja}}}\·V\left(T_0\right)}{(1-\frac{(1-{\mathrm{\η}}_{\mathrm{EQE}})\·I_f}{1/R_{\mathrm{ja}}}\·K_{\mathrm{vc}})}$

Obviously, finding according to above derivation, can be T in junction temperature according to testing sample under laboratory condition ₀time forward voltage V corresponding to LED _f(T ₀), prediction is T at operating ambient temperature _a, forward electric current I=I _fdo the used time, the junction temperature T of LED light source _j, for

I in formula _ffor drive current, T _afor environment temperature, it is known parameters; T ₀represent junction temperature initial value (as environment temperature), V _f(T ₀) expression junction temperature is T ₀time forward voltage, these two parameters can be carried out experiment test acquisition to sample under laboratory condition.η _eQEfor external quantum efficiency, R _jafor the PN junction of LED chip is to the thermal resistance of environment, K _vcfor temperature coefficient, it is unknown parameter to be estimated; Obviously these 3 parameters need to, under laboratory condition, be tested LED sample, and from experimental features curve, adopt curve to carry out parameter extraction, and detailed process is as follows.

First, under experiment condition, while estimating the junction temperature of LED sample, have error by forward voltage method, junction temperature is T ₀time forward voltage V _f(T ₀) relevant with drive current I, therefore, need be according to drive current I=I _fto forward voltage V _f(T ₀) proofread and correct, concrete grammar is as follows,

Obtained by analyzing parameters of semiconductor instrument (Agilent4145) LED sample forward electric current-forward voltage typical test curve (I-V family curve) as shown in Figure 2.

This I-V trial curve also can be described by equivalent physical model, i.e. Shockley diode equation (13),

$I-\frac{(V-{\mathrm{IR}}_s)}{R_p}=I_s\exp \left(\frac{e(V-{\mathrm{IR}}_s)}{n_{\mathrm{ideal}}\mathrm{KT}}\right)---\left(13\right)$

In formula, R _sand R _pbe respectively equivalent parasitic series and parallel connections resistance, n _idealfor ideal factor, I _sfor reverse saturation current, e is elementary charge (1.6 × 10 ^-19coulomb), k is Boltzmann constant (1.38 × 10 ^-23j/K), T represents absolute temperature (unit K).

As V < < V _d=E _gwhen/e, Shockley equation (13) can be approximately,

$I-\frac{V-{\mathrm{IR}}_s}{R_p}\&ap;0---\left(14\right)$

,

$R_p+R_s\&ap;\frac{V}{I}$

If LED quantum hydrazine can be with E _gget 2.9eV, so, the forward voltage of LED is during much smaller than 2.9V, and equation (14) is approximate to be set up.Therefore, can determine equivalent string/parallel resistance (R by the less forward voltage (much smaller than 2.9V) of I-V characteristic test curve and the current ratio of its correspondence _sand R _p) total resistance value; The physical significance of this parameter is: under little electric current (as 100uA) biasing, the forward voltage of LED results from the contribution of connection in series-parallel all-in resistance.

As V > > V _d=E _gwhen/e, Shockley equation (10) is approximately,

$I-\frac{V-{\mathrm{IR}}_s}{R_p}=I_s\exp \left(\frac{e(V-{\mathrm{IR}}_s)}{n_{\mathrm{ideal}}\mathrm{KT}}\right)\&RightArrow;\&infin;$

?

$R_s\&ap;\frac{\mathrm{dV}}{\mathrm{dI}}---\left(15\right)$

Therefore, equivalent string resistance R _scan be similar to and determine at the slope in large voltage range interval (being greater than 2.9V) by I-V characteristic test curve; The physical significance of this parameter is: under large current offset, the forward voltage variable quantity of LED derives from the contribution of equivalent series resistance.

Consider forward electric current I=I in real power illumination _flarge (or forward voltage V _fbe greater than 2.9V), junction temperature is T ₀lED I-V model can be similar to and be reduced to equation (2),

V _f(T ₀)＝I _f·R _s+α ₀(2)

Equivalent string resistance R _swith translational movement α ₀can take method as shown in Figure 2, on the basis of actual I-V attribute testing curve, carry out parameter estimation acquisition according to equation (14) and equation (15).Due to current-crowding effect, cause R _swith electric current I _fincrease and increase, therefore, can adopt electric current I ₀equivalent series resistance R under condition _s(I ₀) revise, be equation (3),

R _s(I _f)＝R _s(I ₀)-α ₁(I _f-I ₀) (3)

Correction factor α ₁by experiment under condition, test drive I=I _fwhen variation, R _s(I) value thereupon changing, works as I=I ₀time, R _s=R _s(I ₀), work as I=I ₁time, R _s=R _s(I ₁), correction factor α ₁for,

${\mathrm{\&alpha;}}_1=\frac{R_s\left(I_1\right)-R_s\left(I_0\right)}{I_1-I_0}---\left(16\right)$

Secondly, K _vcfor temperature coefficient, the temperature coefficient K of LED in practice _vcnot a fixed value, can adopt experiment to calibrate, utilize temperature-forward voltage characteristic (V of the testing of equipment samples such as uA level constant-current circuit, digital multimeter _f-T) as shown in Figure 3, employing formula (4) matching V _f– T slope of a curve, can obtain temperature coefficient K _vcparameter, that is:

$K_{\mathrm{vc}}=\frac{V_f\left(T_j\right)-V_f\left(T_{j0}\right)}{T_j-T_{j0}}---\left(4\right)$

In addition, the PN junction of LED is to the thermal resistance R of environment _janot constant, R _jaincrease with drive current I reduces.According to the R of LED the different forward current I in the situation that _jawith environment temperature T ₀experimental features curve, as shown in Figure 4, establish I ₀thermal resistance when drive current can be measured as R _ja(I ₀), drive current is I _ftime LED thermal resistance R _ja(I _f) can take factor alpha ₂revise, update equation is suc as formula shown in (5).Correction factor α ₂can obtain by family curve shown in Fig. 4 is carried out to matching.

R _ja(I _f)＝R _ja(I ₀)-α ₂(I _f-I ₀) (5)

Adopt exemplary currents-luminous flux trial curve (I-L characteristic) of LED electro-optical investigation system (SSP3112B) the acquisition LED sample based on integrating sphere.N the drive current I[1:N that makes I-L test obtain] (I[k] (k=1,2 ... N)) corresponding Output optical power is φ _l[1:N] (φ [k] (k=1,2 ...), and obtain external quantum efficiency η N) _eQE[k] be,

${\mathrm{\&eta;}}_{\mathrm{EQE}}\left[k\right]=\frac{{\mathrm{\&phi;}}_L\left[k\right]}{I\left[k\right]}---\left(17\right)$

Test and studies show that factually, external quantum efficiency η _eQEwith drive current I _fbetween show typical droop effect, according to LED charge carrier ABC model, external quantum efficiency η _eQEwith drive current I _fbetween relation can be expressed as equation (18),

${\mathrm{\&eta;}}_{\mathrm{EQE}}={\mathrm{\&alpha;}}_3+{\mathrm{\&alpha;}}_4{I}_f+{\mathrm{\&alpha;}}_5{I}_f^2---\left(18\right)$

By I-V characteristic and the I-L characteristic test curve of LED, relation curve (the I-η of sample current-external quantum efficiency of obtaining _eQEcharacteristic) as shown in Figure 5.Factor alpha ₃, α ₄and α ₅can take the family curve of Fig. 5 to carry out curve fitting or peak fitting acquisition.

To sum up, the extraction of LED sample parameters under experiment condition can be realized, the junction temperature prediction under same type LED light sources meaning drive current in office and ambient temperature conditions can be successfully realized by equation (1).Below will be under this predicted condition, carry out the output light flux prediction of LED light source, concrete grammar is as follows.

At junction temperature T ₀time, known according to equation (17), at forward current I _funder output light flux be equation (19),

φ _L(T ₀,I _f)＝η _EQE·I _f(19)

In the time that junction temperature raises, can there is linear reduction in output light flux, so, be T to junction temperature _{j prediction}output light flux proofread and correct and meet equation (7),

φ _l(T _{j prediction}, I _f)=φ _l(T ₀, I _f) (α ₆(T _{j prediction}-T ₀)+α ₇) (7)

Utilize forward electric current-luminous flux characteristic of commercial LED integrated test system SSP3112 test LED sample as shown in Figure 5, factor alpha ₆, α ₇can be obtained through curve by Fig. 6.

In summary, single-chip LED current drives changes following two the main phenomenons of existence:

Phenomenon 1: when drive current rises to I from 0 _ftime, affected by hot loop physical presence thermal capacity, junction temperature rises gradually, reaches equilibrium temperature T _{j prediction}, output light flux L will, through a dynamic changing process, finally reach stable state simultaneously.

Phenomenon 2: in the time that drive current increases gradually, luminous flux increases with current density on the one hand, the forward current constantly increasing on the other hand will cause that junction temperature raises, stop luminous flux to continue to increase, when drive current reaches, certain is suitable when big or small, and now output light flux reaches maximal value.

As can be seen here, at environment temperature T _awith drive current I _funder condition, the junction temperature T of single-chip LED _jand output light flux L can be predicted and be controlled according to method as shown in Figure 1.

Because LED property difference is larger, the junction temperature T of the simplification mathematics physics model of LED to LED _jand the prediction of output light flux and control accuracy greatly depend on experiment test and parameter extraction, main experiment flow and experimental facilities that this invention relates to are as follows:

The basic equipment of LED electro-optic-thermal attribute testing has:

1. for the commercial LED integrated test system SSP3112 of I-V characteristic, I-L characteristic test;

2. for digital multimeter Agilent34401A and the digital oscilloscope TDS3014B of I-T characteristic test;

3. the constant temperature and humidity ageing oven Shicode SDH008 proofreading and correct for H-L characteristic test and temperature coefficient, relative humidity range of control (RH): 0～99%, temperature controlling range (T) :-20 to 200 DEG C;

4. the pulse producer HP8115A producing for PWM drive current;

5. numerical control constant current PWM driving power etc.;

The experiment process concrete steps of LED characteristic research are as follows:

1. preparation of samples

Select LED sample consistent with lighting source, choose at random 20-50, carry out necessary classification.

2. test parameters setting

According to LED sample characteristics of for example, driving parameter is rationally set, is produced the drive current of LED by numerical control constant current PWM driving power; The wet parameter of heat of ageing oven is for being set to respectively: 90 DEG C of relative humidity RH90%, temperature.

3. parameter testing

Test respectively the characteristics such as I-V, I-L, junction temperature, spectrum and the colour temperature of LED; Method for testing junction temperature is: first set constant-current circuit output 100uA electric current, in ageing oven relative humidity, RH is made as 30%, probe temperature is set to even variation within the scope of 30 DEG C to 150 DEG C, its step-length is 10 DEG C, the stable state time of each state is not less than 10 minutes, test GaN forward voltage at each temperature by digital multimeter, the trip temperature coefficient correction of going forward side by side.

4. necessary burn-in test

By the stage, aging test condition being set, is the accelerated aging test of 48 hours to the sample cycle of carrying out in each ageing step, again tests the characteristics such as I-V, I-L, junction temperature, spectrum and the colour temperature of LED.

In actual illumination application, can realize the real-time estimate to LED junction temperature and luminous flux according to above method, ensure under the condition of its life and reliability, by regulating drive current to reach the object of the real-time control of LED junction temperature and luminous flux, application case is as follows.

Based on the control of temperature sensor, its control block diagram as shown in Figure 7.Utilize temperature sensor T _sensorto junction temperature T _jmeasure estimation, according to current drive current I _f, estimate actual output light flux in conjunction with equation (19) and equation (7), then with specifying light flux values φ _setmake comparisons, utilize error to carry out PID and regulate processing, dynamically adjust drive current I _fth.

Without junction temperature prediction and the control of sensor, its control block diagram as shown in Figure 8.Utilize equation (1) to replace the temperature sensor in Fig. 7, estimate drive current I _ftime corresponding junction temperature T _j, estimate actual output light flux in conjunction with equation (19) and equation (7), then with specifying light flux values φ _setmake comparisons, utilize error to carry out PID and regulate processing, dynamically adjust drive current, realize the drive current I without the appointment luminous flux output of sensor _fwocontrol.

Being predicted as example with the junction temperature of three 1W LED chip series mould set verifies.The actual general power of module is 2.64W, and establishing environment temperature is 300K, and so with respect to environment temperature, module 3 LEDs junction temperature of chip are respectively 28.01K, 27.86K and 26.79K; And the result of actual measurement is: module 3 LEDs junction temperature of chip are respectively 30.4497K, 28.74K, and 27.6K, estimated value and measured value coincide and maximal phase is 0.7% to evaluated error.

From above demonstration test, the designed LED junction temperature Forecasting Methodology of the present invention is reliable and effectively, and due to the control structure adopting without sensor, has reduced the usage quantity of system element, improve the reliability of system, there is greatly online detection advantage.

The foregoing is only the preferred embodiments of the present invention, be not limited to the present invention, obviously, those skilled in the art can carry out various changes and modification and not depart from the spirit and scope of the present invention the present invention.Like this, if these amendments of the present invention and within modification belongs to the scope of the claims in the present invention and equivalent technologies thereof, the present invention is also intended to comprise these changes and modification interior.

Claims (7)

1. without prediction and the control method of the LED junction temperature of temperature sensor, it is characterized in that: comprise the following steps S1. by sample experiment test curve, obtain simplification mathematical model and the parameter thereof of LED; S2. utilize the junction temperature T that simplifies mathematical model prediction LED _{j prediction}; S3. according to the junction temperature of predicting in S2, predict the now output light flux φ (T of LED _{j prediction}, I _j); S4. according to the junction temperature of prediction and output light flux pre-conditioned, adjust and upgrade drive current I _f; S5. repeat the process of above prediction and control.

2. prediction and the control method of the LED junction temperature without temperature sensor according to claim 1, is characterized in that: consider environment temperature T _awith drive current I _funder condition, junction temperature T in described step S2 _{j prediction}to predict by equation (1):

Wherein, η _eQEfor external quantum efficiency, R _jafor the PN junction of LED chip is to the thermal resistance of environment, K _vcfor temperature coefficient, T _j0represent initial junction temperature, V _f(T _j0) expression junction temperature is T _j0time forward voltage, I _ffor drive current, T _afor environment temperature.

3. prediction and the control method of the LED junction temperature without temperature sensor according to claim 2, is characterized in that: junction temperature is T _j0time forward voltage V _f(T _j0) I relevant to drive current _f, adopt following methods correction, that is, adopt equation (2) to simplify forward voltage and forward current relationship,

V _f(T _j0)＝I _f·R _s+α ₀(2)

R _sfor equivalent string resistance, α ₀for translational movement, characteristic obtains by experiment,

Due to R _swith electric current I _fincrease and increase, adopt electric current I ₀equivalent series resistance R under condition _s(I ₀) modified R _s,

R _s(I _f)＝R _s(I ₀)-α ₁(I _f-I ₀) (3)

Correction factor α ₁test obtains by experiment.

4. prediction and the control method of the LED junction temperature without temperature sensor according to claim 2, is characterized in that: temperature coefficient K in described equation (1) _vcto obtain by equation (4):

$K_{\mathrm{vc}}=\frac{V_f\left(T_j\right)-V_f\left(T_{j0}\right)}{T_j-T_{j0}}---\left(4\right).$

5. prediction and the control method of the LED junction temperature without temperature sensor according to claim 2, is characterized in that: in described equation (1), junction temperature is to the thermal resistance R of environment _jaobtain by equation (5):

R _ja(I _f)＝R _ja(I ₀)-α ₂(I _f-I ₀) (5)

R _ja(I ₀) record drive current I=I under laboratory condition ₀time LED thermal resistance, R _ja(I _f) be drive current I=I _fthe thermal resistance value of Shi Xiuzheng, correction factor α ₂by the R to LED in different forward current I situations _jawith environment temperature T _afamily curve carry out matching and obtain.

6. prediction and the control method of the LED junction temperature without temperature sensor according to claim 2, is characterized in that: external quantum efficiency η in described equation (1) _eQEobtain by equation (6):

${\mathrm{\&eta;}}_{\mathrm{EQE}}={\mathrm{\&alpha;}}_3+{\mathrm{\&alpha;}}_4{I}_f+{\mathrm{\&alpha;}}_5{I}_f^2---\left(6\right)$

Factor alpha ₃, α ₄and α ₅by the I-V characteristic of LED with I-L characteristic test curve carries out curve fitting or peak fitting obtains.

7. prediction and the control method of the LED junction temperature without temperature sensor according to claim 1, is characterized in that: in described step S3, luminous flux prediction realizes by equation (7)

φ _L(T _j,I _j)＝φ _L(T _j0,I _j)·(α ₆(T _j-T _j0)+α ₇) (7)

I _jexpression junction temperature is T _jtime drive current, factor alpha ₆, α ₇can be obtained by forward electric current-luminous flux characteristic curve fitting.