Abstract

The fault diagnosis and prognosis of wind turbine systems represent a challenging issue, thus justifying the research topics developed in this work with application to safety-critical systems. Therefore, this chapter addresses these research issues and demonstrates viable techniques of fault diagnosis and condition monitoring. To this aim, the design of the so-called fault detector relies on its estimate, which involves data-driven methods, as they result effective methods for managing partial information of the system dynamics, together with errors, model-reality mismatch and disturbance effects. In particular, the considered data-driven strategies use fuzzy systems and neural networks, which are employed to establish non-linear dynamic links between measurements and faults. The selected prototypes are based on non-linear autoregressive with exogenous input descriptions, since they are able to approximate non-linear dynamic functions with arbitrary degree of accuracy. The capabilities of the designed fault diagnosis schemes are verified via a high-fidelity simulator, which describes the normal and the faulty behaviour of a wind turbine plant. Finally, the robustness and the reliability features of the proposed methods are validated in the presence of uncertainty and disturbance implemented in the wind turbine simulator.

Keywords: fault diagnosis, analytical redundancy, fuzzy prototypes, neural networks, diagnostic residuals, fault reconstruction, wind turbine simulator

## 1. Introduction

The increasing level of wind-generated energy in power generation worldwide also increases the levels of reliability and the so-called 'sustainability'shown by wind turbines. Wind turbine systems should generate the required amount of electrical power continuously, depending on the available wind speed, the grid's demand and possible malfunctions.

To achieve this aim, possible malfunctions affecting the process have to be properly detected and managed, before they degrade the nominal working conditions of the plant or become critical issues. Wind turbines with large rotors (i.e., of megawatt size) are very expensive systems, thus requiring an extremely high level of availability and reliability, in order to maximise the generated energy (at a reduced cost), with a minimisation of the operation and maintenance (O&M) services. In fact, the costs of the produced energy are mainly due to the installation cost of the wind turbine, while unplanned O&M costs could increase it up to about 30%, in particular when offshore installations are considered, see Odgaard [1].

These issues have motivated the development of fault diagnosis techniques that can be coupled with the fault-tolerant controllers (the so-called 'sustainable'systems). On the other hand, many turbine manufacturers adopt conservative approaches against faults, which lead to the shutdown of the plant in order to wait for O&M service. Hence, effective tools for coping with faults have to be investigated, in order to improve wind turbine features, particularly during faulty situations. This will lead to prevent critical failures that may affect other wind turbine components, thus avoiding unplanned replacement of functional parts, as well as the decrease of O&M costs, with the increase of the energy production. Moreover, the development of digital control systems, big data tools and artificial intelligence strategies enhance the development of new real-time condition monitoring, diagnosis and fault-tolerant control strategies for industrial processes, which can be available only on demand.

network fault estimators exploits standard training algorithm, that processes the

investigations on more realistic applications of the proposed schemes.

The developed fault diagnosis strategies are verified by means of a high-fidelity simulator, which describes the normal and the faulty behaviour of a wind turbine plant. The achieved performances are verified in the presence of uncertainty and disturbance effects, thus validating the robustness features of the proposed schemes. The effectiveness verified from the achieved results suggests further

The work is organised as follows. Section 2 recalls the wind turbine simulator. Section 3 illustrates the fault diagnosis methodologies relying on fuzzy and neural network prototypes. The obtained results are summarised in Section 4. Finally, Section 5 ends the chapter by outlining the key achievements of the study, and

The wind turbine simulator used in this work was proposed in Odgaard et al. [12]. It describes the realistic behaviour of a three-blade horizontal-axis variablespeed pitch-controlled wind turbine coupled with a full converter generator. The overall system consists of four interconnected modules, that is, the wind driving process, the wind turbine, the measurement system and the baseline controller. The wind turbine block contains three submodels: the blade and the pitch system, the drive-train model and the generator system. The links between the system submodels are represented in Figure 1, with the fault diagnosis system to be designed. The simulator is able to generate several fault scenarios, see

In the following, the description of these interconnected submodels is briefly

data acquired from the process, see Roy and Ganguli [11].

Fault Diagnosis Techniques for a Wind Turbine System DOI: http://dx.doi.org/10.5772/intechopen.83810

providing suggestions for future research issues.

2. Wind turbine simulator description

Odgaard et al. [12].

recalled.

Figure 1.

47

The wind turbine simulator with its fault diagnosis system.

In recent years, many works have been proposed on the topics of fault diagnosis of wind turbines, as shown very recently in Habibi et al. [2] and Lan et al. [3]. Some of them are focused on the diagnosis of particular faults, for example, those affecting the drive-train system at a wind turbine level. Sometimes, these faults are better managed when the wind turbine system is considered in comparison to other parts of the whole plant, see Odgaard & Stoustrup [4]. Moreover, fault-tolerant control of wind turbines has been investigated, for example, in Parker et al. [5] and international cooperations on these problems were also proposed, see Odgaaard and Shafiei [6].

Fault diagnosis oriented to the sustainability feature when applied to safetycritical systems such as wind turbines has been proven to be a challenging issue, see Byrski and Byrski [7] and Xu et al. [8], thus motivating the research topics addressed in this chapter.

This point is fundamental as the increasing demand for energy generation using renewable sources has led to higher attention on renewable energy conversion systems, and in particular wind turbines. They represent very complex and safety-critical plants which require reliability, availability, maintainability and safety. Moreover, their efficiency to generate electrical power has to be maximised. This motivates novel research aspects, in particular in the context of diagnosis and control. The earlier diagnosis of faults and sustainable control solutions can lead to optimise energy conversion and guarantee the desired performances in the presence of possible malfunctions due to unexpected faults and disturbance.

Therefore, this chapter analyses the problem of the fault diagnosis for wind turbine systems, and the development of practical and reliable solutions for fault diagnosis, also known as fault detection and isolation (FDI). Further design of faulttolerant controllers is not considered in this work, but it can rely on the tools considered in this chapter. In fact, the fault diagnosis module provides information on the faulty or fault-free conditions of the system, so that the controller activity can be compensated. This fault diagnosis task is enhanced by the use of fault estimators, which are obtained via data-driven approaches, as they offer effective tools for managing limited analytical knowledge of the process dynamics, together with noise and disturbance effects.

The first data-driven solution considered in this chapter uses fuzzy Takagi-Sugeno models, see Babuška [9], which are derived from a clustering algorithm, followed by an identification procedure, see Simani et al. [10]. A second solution is also considered, which relies on neural networks to describe the non-linear analytical links between measurement and fault signals. The chosen network architecture belongs to the Nonlinear AutoRegressive with eXogenous (NARX) input prototype, which can describe dynamic relationships along time. The training of the neural

#### Fault Diagnosis Techniques for a Wind Turbine System DOI: http://dx.doi.org/10.5772/intechopen.83810

These issues have motivated the development of fault diagnosis techniques that can be coupled with the fault-tolerant controllers (the so-called 'sustainable'systems). On the other hand, many turbine manufacturers adopt conservative approaches against faults, which lead to the shutdown of the plant in order to wait for O&M service. Hence, effective tools for coping with faults have to be investigated, in order to improve wind turbine features, particularly during faulty situations. This will lead to prevent critical failures that may affect other wind turbine components, thus avoiding unplanned replacement of functional parts, as well as the decrease of O&M costs, with the increase of the energy production. Moreover, the development of digital control systems, big data tools and artificial intelligence strategies enhance the development of new real-time condition monitoring, diagnosis and fault-tolerant control strategies for industrial processes, which can be

In recent years, many works have been proposed on the topics of fault diagnosis of wind turbines, as shown very recently in Habibi et al. [2] and Lan et al. [3]. Some of them are focused on the diagnosis of particular faults, for example, those affecting the drive-train system at a wind turbine level. Sometimes, these faults are better managed when the wind turbine system is considered in comparison to other parts of the whole plant, see Odgaard & Stoustrup [4]. Moreover, fault-tolerant control of wind turbines has been investigated, for example, in Parker et al. [5] and international cooperations on these problems were also proposed, see Odgaaard and

Fault diagnosis oriented to the sustainability feature when applied to safetycritical systems such as wind turbines has been proven to be a challenging issue, see

This point is fundamental as the increasing demand for energy generation using renewable sources has led to higher attention on renewable energy conversion

The first data-driven solution considered in this chapter uses fuzzy Takagi-Sugeno models, see Babuška [9], which are derived from a clustering algorithm, followed by an identification procedure, see Simani et al. [10]. A second solution is also considered, which relies on neural networks to describe the non-linear analytical links between measurement and fault signals. The chosen network architecture belongs to the Nonlinear AutoRegressive with eXogenous (NARX) input prototype, which can describe dynamic relationships along time. The training of the neural

Byrski and Byrski [7] and Xu et al. [8], thus motivating the research topics

systems, and in particular wind turbines. They represent very complex and safety-critical plants which require reliability, availability, maintainability and safety. Moreover, their efficiency to generate electrical power has to be maximised. This motivates novel research aspects, in particular in the context of diagnosis and control. The earlier diagnosis of faults and sustainable control solutions can lead to optimise energy conversion and guarantee the desired performances in the presence of possible malfunctions due to unexpected faults and disturbance. Therefore, this chapter analyses the problem of the fault diagnosis for wind turbine systems, and the development of practical and reliable solutions for fault diagnosis, also known as fault detection and isolation (FDI). Further design of faulttolerant controllers is not considered in this work, but it can rely on the tools considered in this chapter. In fact, the fault diagnosis module provides information on the faulty or fault-free conditions of the system, so that the controller activity can be compensated. This fault diagnosis task is enhanced by the use of fault estimators, which are obtained via data-driven approaches, as they offer effective tools for managing limited analytical knowledge of the process dynamics, together

available only on demand.

Fault Detection, Diagnosis and Prognosis

addressed in this chapter.

with noise and disturbance effects.

46

Shafiei [6].

network fault estimators exploits standard training algorithm, that processes the data acquired from the process, see Roy and Ganguli [11].

The developed fault diagnosis strategies are verified by means of a high-fidelity simulator, which describes the normal and the faulty behaviour of a wind turbine plant. The achieved performances are verified in the presence of uncertainty and disturbance effects, thus validating the robustness features of the proposed schemes. The effectiveness verified from the achieved results suggests further investigations on more realistic applications of the proposed schemes.

The work is organised as follows. Section 2 recalls the wind turbine simulator. Section 3 illustrates the fault diagnosis methodologies relying on fuzzy and neural network prototypes. The obtained results are summarised in Section 4. Finally, Section 5 ends the chapter by outlining the key achievements of the study, and providing suggestions for future research issues.

## 2. Wind turbine simulator description

The wind turbine simulator used in this work was proposed in Odgaard et al. [12]. It describes the realistic behaviour of a three-blade horizontal-axis variablespeed pitch-controlled wind turbine coupled with a full converter generator. The overall system consists of four interconnected modules, that is, the wind driving process, the wind turbine, the measurement system and the baseline controller. The wind turbine block contains three submodels: the blade and the pitch system, the drive-train model and the generator system. The links between the system submodels are represented in Figure 1, with the fault diagnosis system to be designed. The simulator is able to generate several fault scenarios, see Odgaard et al. [12].

In the following, the description of these interconnected submodels is briefly recalled.

Figure 1. The wind turbine simulator with its fault diagnosis system.

#### 2.1 Wind turbine model

The turbine system consists of three submodels motivated by the power transmission flow. First, the blade and pitch block represents how the blades capture wind energy, which is based on the following aerodynamic law:

$$\tau\_r(t) = \frac{\rho \,\pi R^3 \,\mathrm{C\_q}(\lambda(t), \beta(t)) v\_w^2(t)}{2} \tag{1}$$

simulate the real behaviour of sensors and actuators. Therefore, the measured signals are modelled as sum of their actual value and white Gaussian process terms. Moreover, the wind turbine simulator includes a baseline controller, represented by a PID standard regulator, which regulates the generated power on the basis of the actual wind speed, as shown in Odgaard & Stoustrup [4] and Odgaard et al. [12].

The wind turbine simulator includes the generation of three different typical fault cases, that is, sensor, actuator and system faults, see Odgaard and Stoustrup

For the case of the sensor faults, they are generated as additive signals on the affected measurements. As an example, the faulty sensor of faulty pitch angle β<sup>m</sup> provides wrong measurements on blade orientation; thus, if not handled, the con-

On the other hand, actuator faults lead to the alteration of pitch angle or the generator torque transfer functions of Eqs. (2) and (4), by modifying their dynamics. They simulate a pressure drop in the hydraulic circuit of the pitch actuator or an

Finally, a system fault affects the drive train of the turbine, which is described as a slow variation in time of the friction coefficient. This can be due to the effect of

These nine fault cases are summarised in Table 1, which also highlights which

With these assumptions, the overall model of the wind turbine process can be represented as a non-linear continuous-time function fwt describing the evolution

> x\_ wtðÞ¼ t fwt xwt ð Þ ; uð Þt yðÞ¼ t xwtð Þt

where in this case, the state of the system is considered equal to the monitored

xwtðÞ¼ t yðÞ¼ t ωg,m1;ωg,m2;ωr,m1;ωr,m2; Pg,m

Fault case Fault type Affected measurement

 Sensor β1,m<sup>1</sup> Sensor β2,m<sup>2</sup> Sensor β3,m<sup>1</sup> Sensor ωr,m<sup>1</sup>

8 Actuator τg,m 9 System Drive train

5 Sensor ωr,m<sup>2</sup> and ωg,m<sup>2</sup>

6 Actuator Pitch system of blade #2 7 Actuator Pitch system of blade #3

system output, that is, the rotor speed, the generator speed and the generated

(6)

2.2 Simulated fault scenario

[4] and Odgaard et al. [12].

power:

Table 1.

49

Wind turbine simulator fault scenario.

troller cannot fully track the power reference signal.

electronic break down in the converter device.

Fault Diagnosis Techniques for a Wind Turbine System DOI: http://dx.doi.org/10.5772/intechopen.83810

wear and tear along time of the mechanical parts.

measured signals are affected by them, as shown in Figure 1.

of the turbine state vector xwt excited by the input vector u:

For each blade, Eq. (1) describes the torque acting on the rotor τr, depending on the squared wind speed v<sup>2</sup> <sup>w</sup>, the air density ρ and the rotor radius R. The coefficient Cq is usually defined using a two-dimensional map depending on the blade pitch angle β and the tip-speed ratio λ, that is, the ratio between the linear velocity of the blade tip and the wind speed. This map is represented by means of a look-up table. The blade and pitch system includes the dynamics of the pitch angle hydraulic piston servo system, which is approximated as a second-order transfer function of Eq. (2):

$$\frac{\beta(s)}{\beta\_{r\not\!f}(s)} = \frac{o\_n^2}{s^2 + 2\zeta\alpha\_n s + o\_n^2} \tag{2}$$

where βref is the reference pitch angle computed by the turbine controller, while ζ and ω<sup>n</sup> are the transfer function parameters.

The drive-train system determines the power flow through the gear box from the rotor toward the electric generator, whose dynamics are described as in Eq. (3):

$$\begin{cases} J\_r \dot{\boldsymbol{\alpha}}\_r = \boldsymbol{\tau}\_r - K\_{dt} \boldsymbol{\theta}\_\Delta - (B\_{dt} + B\_r) \boldsymbol{\alpha}\_r + \frac{B\_{dt}}{N\_\mathcal{g}} \boldsymbol{\alpha}\_\mathcal{g} \\\\ J\_\mathcal{g} \dot{\boldsymbol{\alpha}}\_\mathcal{g} = \frac{\eta\_{dt} K\_{dt}}{N\_\mathcal{g}} \boldsymbol{\theta}\_\Delta + \frac{\eta\_{dt} B\_{dt}}{N\_\mathcal{g}} \boldsymbol{\alpha}\_r - \left(\frac{\eta\_{dt} B\_{dt}}{N\_\mathcal{g}^2} + B\_\mathcal{g}\right) \boldsymbol{\alpha}\_\mathcal{g} - \boldsymbol{\tau}\_\mathcal{g} \\\\ \dot{\boldsymbol{\theta}}\_\Delta = \boldsymbol{\alpha}\_r - \frac{\boldsymbol{\alpha}\_\mathcal{g}}{N\_\mathcal{g}} \end{cases} \tag{3}$$

where Jr and Jg are the inertia moments of the rotor and generator shafts, respectively. Kdt is the torsion stiffness, Bdt is the torsion damping factor, Bg is the viscous friction of the generator shaft, Br is the viscous friction of the low-speed shaft, Ng is the gear ratio, ηdt is the efficiency and θ<sup>Δ</sup> is the torsion angle.

Finally, the generator submodel represents the converter dynamics by means of first-order transfer function of Eq. (4):

$$\frac{\tau\_{\mathfrak{g}}(s)}{\tau\_{\mathfrak{g},r\mathfrak{g}'}(s)} = \frac{a\_{\mathfrak{g}}}{s + a\_{\mathfrak{g}}} \tag{4}$$

where τg,ref is the reference torque defined by the controller and α<sup>g</sup> is the transfer function parameter.

Finally, the generated power Pg is computed as the product of the generator torque by its speed, decreased by the efficiency coefficient ηg:

$$P\_{\mathfrak{g}} = \eta\_{\mathfrak{g}} \, o\_{\mathfrak{g}} \, \mathfrak{r}\_{\mathfrak{g}} \tag{5}$$

As sketched in Figure 1, the signals generated by the wind turbine system are assumed to be acquired through the measurement block, whose objective is to

simulate the real behaviour of sensors and actuators. Therefore, the measured signals are modelled as sum of their actual value and white Gaussian process terms. Moreover, the wind turbine simulator includes a baseline controller, represented by a PID standard regulator, which regulates the generated power on the basis of the actual wind speed, as shown in Odgaard & Stoustrup [4] and Odgaard et al. [12].
