【正文】 s, model reference adaptive systems and use of multivariable control structure. Most of these controllers use mathematical models and are sensitive to parametric variations. Very few adaptive controllers have been practically employed in the control of electric drives due to their plexity and inferior performance.The design of current and speed controllers for permanent magnet brushless DC(PMBLDC) motor drive remains to large extent a mystery in the motor drives field. A precise speed control of PMBLDC motor is plex due to nonlinear coupling between winding currents and rotor speed as well as nonlinearity present in the developed torque due to magnetic saturation of the rotor.The PMBLDC machines can be categorized based on the permanent magnets mounting and shape of the backEMF. The permanent magnets can be surface mounted on the rotor or installed inside of the rotor (interior permanent magnet), and the backEMF shape can either be sinusoidal or trapezoidal. The surface mounted PM (SMPM) machine is easy to build. Also, from the machine design point of view, skewed poles can be easily magnetized on this round rotor to minimize cogging torque. Typically, for this type of motor, the inductance variation by rotor position is negligibly small since there is no magnetic saliency. The interior permanent magnet (IPM) machine is a good candidate for highspeed and traction applications. It is noted that there is an inductance variation by rotor position for this type of motor because of the magnetic saliency.This paper will give bigger focus on the artificial intelligent applications to PMBLDC motor drives. In this paper, conventional and recent advancement of AI operation methods for P M BLDC drives are presented.2 Modelling of PMBLDC motorThe PMBLDC motor is modelled in the stationary reference frame using 3 phase abc variables (Pillay and Krishnan 1989). The general voltampere equation be expressed as:where R , L , M are the resistance, inductance and mutual inductance of stator windings and ， are phase voltage, backEMF voltage and phase current of each phase of stator respectively. The electromagnetic torque is expressed asFig. 1 Three phase back EMF functionThe interaction of with the load torque determines how the motor speed builds up: where is load torque in N m, B is the frictional coefficient in N ms/ rad, and J is the moment of inertia, kg㎡. The per phase back emf in the PMBLDC motor is trapezoidal in nature and are the functions of the speed and rotor position angle (θ r). The normalized functions of back emfs are shown in Fig. 1. From this, the phase back emf can be expressed as: Where and can be described by E and normalized back emf function shown in Fig. 1. . The back emf function of other two phases and are defined in similar way using E and the normalized back emf function and as shown in F ig. 1.3 .Artificial intelligenceHuman abilities in controlling the plex systems, has encouraged scientists to pattern from human neural network and decision making systems. Firstly there searches began in two separate fields and resulted in establishment of the fuzzy systems and artificial neural networks (Giridharan e t a l. 2006). There are primarily three concepts prevailing over the intelligent control:? Fuzzy logic control? Neural network based control? Neuro fuzzy control (hybrid control)In the first concept, the controller is represented as a set of rules, which accepts/gives the inputs/outputs in the form of linguistic variables. The main advantages of such a controller are:Fig. 2 PMBLDC motor AI controllers scheme(1) Approximate knowledge of plant is required(2) Knowledge representation and inference is simple.(3) Implementation is fairly easy.The artificial intelligence mainly has two functions in PMBLDC motor drivesa. Artificial intelligence control—As controllerb. Sensorless operations—for variable estimationIn these the conventional controllers like PI，PID etc. are replaced or bined with AI controllers. All artificialintelligencebased control strategies, such as fuzzy logic control, neural network control, neurofuzzy control, and genetic control, are classified as artificial intelligent control (AIC). Among them, the fuzzy logic control and the neural network control are most mature and attractive for the PMBLDC drives since they can effectively handle the system’s nonlinearities and sensitivities to parameter variations (Fig. 2).附錄C 中文譯文基于人工智能在永磁無刷直流電機(jī)驅(qū)動(dòng)中的應(yīng)用摘要由于其結(jié)構(gòu)簡(jiǎn)單和低成本的原因，永磁無刷直流電機(jī)越來越受到青睞。永磁材料的進(jìn)步和電力電子器件的可靠性能，使得永磁無刷直流電機(jī)成本更加低廉，有了更多方面的應(yīng)用。人工智能應(yīng)用的進(jìn)步如神經(jīng)網(wǎng)絡(luò)算法、模糊邏輯算法、遺傳算法等對(duì)電機(jī)驅(qū)動(dòng)器產(chǎn)生了巨大的影響。無刷直流電機(jī)是一個(gè)多變量、非線性系統(tǒng)。在傳統(tǒng)的永磁無刷直流電機(jī)中驅(qū)動(dòng)器的速度和在無刷直流電機(jī)中的位置檢測(cè)都需要很高的精度。不幸的是,傳統(tǒng)的控制方法需要詳細(xì)的所有電動(dòng)機(jī)參數(shù)建模來實(shí)現(xiàn)這個(gè)。智能控制技術(shù)、模糊邏輯控制、神經(jīng)網(wǎng)絡(luò)控制等使用啟發(fā)式的輸入輸出關(guān)系來處理模糊和復(fù)雜的情況。本文是一篇將智能控制技術(shù)用于永磁無刷直流電機(jī)驅(qū)動(dòng)的調(diào)查文獻(xiàn)。各種人工智能技術(shù)對(duì)永磁無刷直流電機(jī)驅(qū)動(dòng)描述。試著為那些工作在永磁無刷直流電機(jī)領(lǐng)域的執(zhí)行工程師和研究員們提供一種指導(dǎo)和有效地參考。關(guān)鍵詞：永磁無刷直流電機(jī)，人工智能，智能控制，模糊控制，神經(jīng)網(wǎng)絡(luò) 永磁無刷直流電機(jī)越來越多的被應(yīng)用于各種應(yīng)用場(chǎng)合并且它的市場(chǎng)也在快速地增長(zhǎng)。這主要是由于其高轉(zhuǎn)矩、緊密度、效率高。永磁無刷直流電機(jī)由于其功率密度大并易于控制被廣泛投入到使用中。高性能永磁材料和電力電子器件的進(jìn)步已經(jīng)廣泛的應(yīng)用在提高永磁無刷直流電機(jī)中的變速驅(qū)動(dòng)器中類似于交流電機(jī)(辛格和庫馬爾2002。玻色1992)。最近,永磁無刷直流電動(dòng)機(jī)已逐漸成為替代標(biāo)準(zhǔn)的刷式直流電機(jī)，由于其效率高、維護(hù)成本低和良好的可控性(Mohan et al . 1995) 在許多伺服應(yīng)用中都被用到。已經(jīng)提出幾個(gè)模型的驅(qū)動(dòng)并為之討論(PuttaSwamy e t1995)。此外,永磁無刷直流電機(jī)是一種同步電機(jī)，也就是說由定子和轉(zhuǎn)子所產(chǎn)生的磁場(chǎng)都有相同的頻率。因此, 永磁無刷直流電機(jī)不會(huì)出現(xiàn)通常出現(xiàn)在感應(yīng)電動(dòng)機(jī) “滑”的這種現(xiàn)象, (Hendershot和米伊勒河1994)。該項(xiàng)研究是為永磁無刷直流電機(jī)確定一個(gè)合適的速度控制器。在經(jīng)典的線性理論中人們提出了許多控制策略 (2001。1989年Kaynak Miller)。隨著永磁無刷直流電機(jī)作為非線性模型、線性PID可能不再適和當(dāng)前的對(duì)象。這已經(jīng)導(dǎo)致現(xiàn)代非線性自整定控制器，狀態(tài)反饋控制器，模型參考自適應(yīng)系統(tǒng)和多變量控制結(jié)構(gòu)中使用的入耳式控制結(jié)構(gòu)需求的增加。大多數(shù)這些控制器使用的數(shù)學(xué)模型和運(yùn)行參數(shù)的變化都很敏感。由于其復(fù)雜的密度和低質(zhì)的性能幾乎很少的自適應(yīng)控制器用于控制電動(dòng)驅(qū)動(dòng)器。永磁無刷直流電（PMBLDC）機(jī)驅(qū)動(dòng)電流及速度控制器的設(shè)計(jì)上在電機(jī)驅(qū)動(dòng)領(lǐng)域中很大程度上仍然是一個(gè)謎。對(duì)于永磁無刷直流電機(jī)來說精確的速度控制是復(fù)雜的，由于繞組之間的電流和轉(zhuǎn)子速度，以及由于磁飽和轉(zhuǎn)子的轉(zhuǎn)矩中存在非線性的非線性耦合。永磁無刷直流電機(jī)的分類可以基于永磁體安裝和反電動(dòng)勢(shì)的波形來區(qū)分。永久磁鐵可以安裝在轉(zhuǎn)子的表面或安裝在轉(zhuǎn)子的內(nèi)部(內(nèi)置式永磁)，反電動(dòng)勢(shì)的形狀可以是正弦波或是梯形波。表面安裝PM(SMPM)機(jī)器很容易建立。同時(shí),從機(jī)械設(shè)計(jì)的角度來看,傾斜磁極可以很容易在圓形磁極上磁化，以盡量減少齒槽轉(zhuǎn)矩。通常,對(duì)于這種類型的電動(dòng)機(jī)而言, 由轉(zhuǎn)子位置的的電感距離變化，是小到可以忽略不計(jì)，因?yàn)闆]有磁性的顯著性。內(nèi)部永磁（IPM）的機(jī)器，是高速和牽引應(yīng)用的一個(gè)很好的備用。值得注意的是，因?yàn)檫@種類型的磁凸極轉(zhuǎn)子位置的變化它又是電感電機(jī)。本文將會(huì)在人工