Adaptive dynamic programming (ADP) is a sub-field of approximate dynamic programming that deals with the adaptive control of continuous nonlinear dynamic systems. Its origins stem from dynamic programming in optimal control, but it is extended into a form where approximations are used to reduce the curse of dimensionality and reduce the need for model knowledge. ADP is also considered to be one of the main reinforcement learning (RL) approaches since it uses information obtained from interaction with the environment to improve its policy. RL in general and ADP in particular are well suited for application to autonomous aerospace systems, since they allow adaptive control in case of uncertainties or faults in the system, even if the fault is of a type that is not anticipated during the control design. This chapter first gives a brief historical overview of ADP applications to flight control tasks. After that, four recent advances of ADP for flight control are presented.

@en

In this paper, a physics of failure-based prediction method is combined with statistical models to consider the impact of current crowding and current droop effects on the reliability of LED arrays. Electronic-thermal models of LEDs are utilized to obtain the operation conditions under the influences of current crowding and current droop. A Markov chain-based model is used to calculate the probability distribution of each failure mode, including the lumen decay and catastrophic failure. Two types of LEDs were selected for a numerical study. The proposed prediction method provides the realistic reliability prediction results. It is found that the properties of LEDs have a great impact on their hazard rates of LED arrays. The equivalent resistance, third-order non-radiative coefficient, and radiative coefficient of LEDs are critical to the reliability of an LED array.

@en

Nowadays, due to the advancement of design and manufacturing technology, there are many consumer products with high reliability. Similarly, the competition in the business sector influences the product development time to become shorter and that makes it difficult for manufacturers to evaluate the reliability of current products before new products are released to the market. This phenomenon is manifested in the lighting industry, especially for the high power white light-emitting diodes (LEDs) as these products have a long lifetime and high reliability. Currently, the standard to predict the lifetime of LEDs is based on a deterministic nonlinear least squares method which has low prediction accuracy. To overcome this, degradation models are being used to study the reliability of such products, considering the uncertainties and the quality characteristics whose degradation over a period of time can be related to the product lifetime. A stochastic approach based on gamma distributed degradation (GDD) is proposed in this study to estimate the long-term lumen degradation lifetime of phosphor-converted white LEDs. An accelerated thermal degradation test was designed to gather luminous flux degradation data which was analyzed based on maximum likelihood estimation (MLE) and the method of moments (MM) to estimate the parameters for the GDD model. The MLE method has shown superiority over MM in terms of the estimation of the model parameters due to its iterative algorithm being likely to find the optimal estimation. The lifetime prediction results show that the accuracy of the proposed method is much better than the TM-21 nonlinear least squares (NLS) approach which makes it promising for future industrial applications.

@en

This work studies the effect of randomness of LED's lumen depreciation on reliability of the entire LED lamp. An integrated LED light bulb is selected as carrier of the proposed method. A PoF based lumen depreciation model and electronic-thermal simulations are introduced for reliability prediction. The normal distribution is used to describe the statistical distribution of LEDs. The probabilities of the driver's catastrophic failures and lumen can then be obtained by Monte Carlo simulations by considering the increase of lamp's temperature. The effect of the lumen depreciation to the entire lamp is studied with two scenarios: constant light mode and constant current mode.

@en

As an increasing attention towards sustainable development of energy and environment, the power electronics (PEs) are gaining more and more attraction on various energy systems. The insulated gate bipolar transistor (IGBT), as one of the PEs with numerous advantages and potentials for development of higher voltage and current ratings, has been used in a board range of applications. However, the continuing miniaturization and rapid increasing power ratings of IGBTs have remarkable high heat flux, which requires complex thermal management. In this paper, studies of the thermal management on IGBTs are generally reviewed including analyzing, comparing, and classifying the results originating from these researches. The thermal models to accurately calculate the dynamic heat dissipation are divided into analytical models, numerical models, and thermal network models, respectively. The thermal resistances of current IGBT modules are also studied. According to the current products on a number of IGBTs, we observe that the junction-to-case thermal resistance generally decreases inversely in terms of the total thermal power. In addition, the cooling solutions of IGBTs are reviewed and the performance of the various solutions are studied and compared. At last, we have proposed a quick and efficient evaluation judgment for the thermal management of the IGBTs depended on the requirements on the junction-to-case thermal resistance and equivalent heat transfer coefficient of the test samples.@en

In this work, a physics-of-failure (PoF) reliability prediction methodology is combined with statistical models to consider the interaction between the lumen depreciation and catastrophic failures of LEDs. The current in each LED may redistribute when the catastrophic failure occurs in one of LEDs in an array, thus affecting the operation conditions of the entire LED array. A physics-of-failure based reliability prediction methodology is combined with statistical models to consider the interaction between the lumen depreciation and the catastrophic failure. Electronic-thermal simulations are utilized to obtain operation conditions, including temperature and current. Meanwhile, statistical models are applied to calculate possibilities of the catastrophic failure in different operation conditions.

@en

This paper studies the interaction of catastrophic failure of the driver and LED luminous flux decay for an integrated LED lamp with an electrolytic capacitor-free LED driver. Electronic thermal simulations are utilized to obtain the lamp's dynamic history of temperature and current for two distinct operation modes: constant current mode (CCM) and constant light output (CLO) mode, respectively. Driver's mean time to failure (MTTF) and the LED's lifetime in terms of luminous flux are calculated. Under CLO mode, the LED's current increases exponentially to maintain the constant light output. As a result, the junction temperatures of LEDs, MOSFETs, and power diodes in driver rise significantly, leading to a much shorter MTTF and faster luminous flux depreciation. However, under the CCM, the junction temperatures of LEDs, MOSFETs, and diodes change modestly; therefore, the driver's MTTF and LED's luminous flux decay are not affected much by the variation of temperatures during LED's degradation process.

@en

In this paper, an integrated LED lamp with an electrolytic capacitor-free driver is considered to study the coupling effects of both LED and driver's degradations on lamp's lifetime. An electrolytic capacitor-less buck-boost driver is used. The physics of failure (PoF) based electronic thermal simulation is carried out to simulate the lamp's lifetime in three different scenarios: Scenario 1 considers LED degradation only, Scenario 2 considers the driver degradation only, and Scenario 3 considers both degradations from LED and driver simultaneously. When these two degradations are both considered, the lamp's lifetime is reduced by about 22% compared to the initial target of 25,000 h. The results of Scenario 1 and 3 are close to each other. Scenario 2 gives erroneous results in terms of luminous flux as the LED's degradation over time is not taken into consideration. This implies that LED's degradation must be taken into considerations when LED and driver's lifetimes are comparable.

@en

Light-Emitting Diodes (LEDs) have become a very promising alternative lighting source with the main advantages of a longer lifetime and a higher efficiency than traditional ones. However, the LED lamp’s lifetime is compromised by its driver’s reliability. Although extensive studies have been made on the reliability of LEDs, the research on the lifetime prediction for LED drivers, and the interaction of the reliability between LEDs and driver in an LED system is still lacking. This dissertation investigates the lifetime predictions for LED drivers and LED lamps using physics of failure (PoF) based reliability simulations. Various reliability and statistical methods, such as theMonte Carlo method, the fault-tree method, the Markov Chain method, and the Wiener process, are applied and integrated with the electronic-thermal simulation in order to investigate various problems.@en

The temperature of electrolytic capacitor in light-emitting diode (LED) drivers continuously increases under operation conditions, thus the capacitors degrade faster than that with constant temperature assumption. In this paper, a physics-of-failure (PoF)-based reliability prediction methodology is developed for LED drivers to consider the temperature change of electrolytic capacitor. SPICE simulation, compact thermal modeling, and Monte Carlo simulation are integrated to predict the failure rate distribution of an electrolytic capacitor of given LED driver systems. The simulation results agree well with the accelerated test results for an RC linear AC-DC converter. Furthermore, a single inductor buck-boost DC-DC converter is simulated to understand the degradation behavior of electrolytic capacitor. It has been found that the temperature of an output stage capacitor increases significantly during operation time. The capacitor's performance without taking temperature change into account results in an overestimated driver lifetime by more than 38% for the selected case study.@en

Lumen decay of LEDs is affected by time, junction temperature and input current. In LED lamps, both temperature and input current vary with time due to driver's degradation and temperature change of lamps. This paper proposes a lumen decay prediction method which considers effects of operation time, temperature and current by taking the interaction of LED and driver performance into account. In particular, a lumen decay model for LED source with consideration of the ever changing current and temperature is developed and validated experimentally. Several scenarios are analyzed theoretically by applying different assumptions.@en

In this paper, a systematic approach for reliability assessment of electrolytic capacitor-free LED drivers is developed to investigate the failure rate of MOSFETs in the drivers. An original fly-back driver with electrolytic capacitor and two modified electrolytic capacitor-free drivers are studied, and their performances are compared. Due to LED lumen depreciation during operation, current of the MOSFET decreases over time, which cancels out the effect of junction temperature increase. As a result, the junction temperature of MOSFET increase mildly for the case study, but failure rate is accelerated.@en