Introduction to Hydraulic Components 221

Introduction to Hydraulic Components provides users with an overview of how the active and passive components of a hydraulic system work together to transmit power. The active components of a hydraulic system are the hydraulic pump, control valves, and the actuator. Fluid conductors and fluid storage containers are passive components. Each part of a hydraulic system contributes to the manipulation of pressurized hydraulic fluid in order for the system to perform work.

After completing Introduction to Hydraulic Components, users will have an understanding of how the main components of a hydraulic system work together to convert hydraulic energy into mechanical power. Fluid system operators should be knowledgeable about the functions of hydraulic system components and how each part contributes to the success of the hydraulic system.

Huade Hydraulic contains other products and information you need, so please check it out.

• Difficulty Intermediate

• Format Online

• Number of Lessons 19

• Language English

4.1

Communication and Human&#;Machine Interaction

Modern factories need to build a network of sensing, measurement, and control, which covers the whole production process. Therefore, as an important control component of modern factories, the electro-hydraulic valve also needs to have a measurement function for the pressure, flow, and other signals. Meanwhile, the control function is necessary for quick debugging and integration according to the working conditions, as well as a quick information exchange function for communication through a bus or Ethernet [85].

4.1.1

Fieldbus-Based Communication Modes

Fieldbus technology makes the hydraulic valve an independent control element that can achieve decentralized control and centralized management in the industrial field. At the same time, the hydraulic valve and the hydraulic cylinder can form a decentralized independent axis controller by a Fieldbus. The Fieldbus connects the communication unit, which is composed of the electro-hydraulic valve and its pressure, flow, and position sensors, with other hydraulic units, electronic units, and central industrial control units to form a data interaction network. This can achieve control of the entire hydraulic system and the collection of basic information.

There are many studies on integrated communication technology for the electro-hydraulic valve. In a study by Keuper et al., there is a hydraulic proportional valve with an embedded electronic controller. This is applied in agriculture, which achieves the transmission of position data for the valve spool through the controller area network bus (CAN bus) [86]. Tapio Virvalo et al. of the Tampere University of Technology in Finland applied the CAN bus technology into a proportional servo valve. Controlling the communication structure of the system connected by the CAN bus is illustrated in Figure 14. An intelligent two-wire valve based on the power line carrier and the CEBUS Fieldbus has been designed, simulated, and tested by Li Weibo of Zhejiang University. This verified the feasibility and the advantages of Fieldbus technology in the field of hydraulic valves [87].

Communication structure of the control system connected by the CAN bus

Full size image

With the process of digitalization in the hydraulic industry, there is a growing demand for a remote monitoring capability of electro-hydraulic valves. Many enterprises have released a series of hydraulic valves with a communication interface. Most of those valves equip Fieldbus interfaces since they have a huge application market and mature hardware technology. ATOS DLHZO-TEZ series valves and MOOG D636 provide CAN, Profibus DP, EtherCAT interface for users to choose from. HYDAC P4WERE06 integrates the LIN interface. Parker D1FC integrates EtherCAT. Some other control blocks, like HAWE PSL series, Danfoss PVG series, Eaton CMA200, integrate the CAN bus interface.

4.1.2

Bluetooth-Based Communication Modes

In some industrial scenes, where the electro-hydraulic valves do not have enough place for communication cables, it is necessary to make the communication networks wireless. On the other hand, wireless communication networks can make it convenient for engineers to debug these valves and to monitor the equipment states. In regard to the demand for wireless communication networks, some enterprises have released products that are integrated with Bluetooth. SUN A24, a proportional cartridge valve, integrates Bluetooth. HAWE RV2S-BT integrates Bluetooth, which makes valves convenient and intelligent to control, monitor and manage. Figure 15 shows what wireless communication can do.

Function of wireless communication

Full size image

4.1.3

IO-Link-Based Communication Modes

There are lots of Fieldbus standards in different industries, which makes it a huge challenge for users to connect with sensors, actuators, and valves with different communication interfaces in a project. A communication interface called IO-Link can solve this problem. IO-Link is an open standard so that products that are integrated with IO-Link can almost connect with any equipment with a different Fieldbus. This advantage makes it easier for these products to combine with other pieces of equipment that has an independent control system. However, it can only achieve point-to-point communication, and it cannot combine multiple devices into one communication network without the IO-Link Master or the IO-Link Hub. Rexroth 4WRPEH is a product that is integrated with IO-Link.

4.1.4

Human&#;Machine Interaction Software

As a hardware basis, those interfaces can build a huge communication network. However, what makes it incredible is the software behind the network. We can control, monitor, and manage those valves remotely and in real-time using PC software, such as ATOS Z-SW, Parker ProPxD, MOOG Valve, and Pump Configuration Software. At the same time, some valve faults, such as communication errors, hardware errors, and overloading the overloading the power supply voltage, can be diagnosed remotely by that software. Some other applications are designed for those mobile devices, such as SUN AmpSet Blue and HAWE eControl.

4.2

Intelligent Fault Diagnosis

Equipment must be designed so that it is safe, reliable, and of high quality [88]. The EU issued and implemented the Machinery Safety Directive /42/EC. With the development of statistical mathematical tools, the reliability and life of hydraulic products can be calculated quantitatively by using mathematical tools. In addition, the evaluation criteria are more specific. Internationally famous manufacturers have started to carry out product reliability tests and fault diagnosis. Most products in the current market focus on diagnosing some simple faults for electro-hydraulic valves, such as electrical and communication failures. In the application scenario, valves often meet some internal deep faults, which cannot be detected by an existing sensor network.

4.2.1

Common Faults of Electro-Hydraulic Valves

Marco Münchhof built a specialized experiment rig for faults diagnosis for a valve-controlled cylinder system [89]. Figure 16 shows the result of his experiment and which faults will often occur in different elements of the valve-controlled cylinder system. The proportion of each component fault in the system is shown in Figure 17(a). As can be seen from Figure 17(a), in the valve-controlled cylinder system, the fault of the valve spool and the valve body accounts for 51%. Therefore, Münchhof further summarized the fault of the valve body and the valve spool as illustrated in Figure 17(b) and (c).

Valve control cylinder system faults and the corresponding components

Full size image

(a) The proportion of each element failure; (b) The proportion of the valve body faults; (c) The proportion of the valve spool faults

Full size image

Denson Hydraulic has provided the literature with some common failures for the proportional directional valve in engineering practice and their causes as shown in Figure 18 [90].

Common valve faults and their causes

Full size image

From these figures above, a conclusion can be drawn that many faults occur in engineering practice because of the system design and personal operation errors. At the same time, we cannot monitor the movement state and the system operation state of the valve without suitable measuring devices so that the fault can only be detected when the fault is obvious. Hence, studies on intelligent fault diagnosis of the valve are vital to find the fault in time and to prevent the system from losing control. Therefore, the focus of fault diagnosis is how to extract the fault features from multi-sensor signals and to judge the abnormal state. According to the theory and methods of fault diagnosis, the common methods in the field of the hydraulic systems can be divided into two categories: a data-based diagnosis method and a model-based diagnosis method [91]. Figure 19 shows the common methods in the fault diagnosis of the electro-hydraulic valve.

Common methods in fault diagnosis of electro-hydraulic valve

Contact us to discuss your requirements of Hydraulic pressure valve. Our experienced sales team can help you identify the options that best suit your needs.

Full size image

4.2.2

Data-Based Fault Diagnosis Methods

The data-based fault diagnosis method mines the hidden information in the data through various data analysis methods to predict the faults; however, the acquisition cost of vast data is high, and the uncertainty and incompleteness of the data will affect the accuracy of this method. In terms of practical applications, it is in real-time and it is a practical method to detect a signal threshold to diagnose the faults. Some electro-hydraulic valve products, like Eaton AxisPro, install sensors in the valve to improve the state acquisition ability, and it also uses the numerical threshold to monitor the operation states of the valve. Afterwards, the state quantities of the valve are checked to see if they exceed the preset threshold value to diagnose the faults.

This method has been extended by Raduenz et al. According to a large number of collected data, the threshold range of the fault diagnosis has been effectively reduced [92, 93]. For a specific proportional valve, the relationship between the control signal and the displacement of the valve spool changes in a certain range with the change of the working conditions. If the fault occurs, the corresponding relationship will change. When the same control signal is given, the actuator or valve spool displacement will exceed the specified threshold value, which can be used for fault diagnosis.

Because the electro-hydraulic valve usually works in a relatively stable state and there is less effective high-frequency information in the signal, many researchers have chosen the signal during the step response of the valve spool as the reference signal. For example, Sharifi extracted the step amplitude, the peak generation time, and the oscillation time in the pressure step curve generated by the system during the step process of the valve spool. The features under the normal working conditions and the leakage working conditions were analyzed in clusters through the neural network to identify the leakage fault, as depicted in Figure 20 [94, 95]. In addition, the error of the fault identification is 30.2%.

Fault feature selection and the clustering schematic diagram [94, 95]

Full size image

In order to increase the fault information and improve the sensitivity of the fault diagnosis, Zhao et al. [96] and Chao [97] adopted the method of wavelet packet signal decomposition to process the pressure signal and use the energy distribution for each level of the wavelet to diagnose the leakage fault. Wang et al. [98] further combined the wavelet packet signal decomposition method with the feature data principal component analysis method while adding the fault features with the feature data extracted from the characteristic curve to facilitate rapid fault diagnosis.

The lack of sensor information still severely limits the fault diagnosis ability for the method mentioned above. Therefore, on the basis of adding sensors, the state information collected by the pressure sensor and displacement sensor was integrated to establish the corresponding relationship between the fault phenomenon and the fault signal. In addition, the fault was diagnosed by calculating the membership degree [99]. At the same time, many researchers [100,101,102,103,104,105] have used the neural network or similar methods to link the valve characteristic curve and fault in case of failure to avoid the problem of manually finding the signal characteristics.

However, the above methods can easily be affected by the characteristics of the training data and they need a large amount of data to support the training process. To make up for the above shortcomings, some researchers use the neural network to establish the hydraulic system model and take the calculated deviation as the fault diagnosis basis [106]. This method reduces the requirement of the training data and the fault features contain more physical information, which is easy to understand and further modify. For example, Liu et al. [107] proposed a two-stage RBF (radial basis function) neural network as the fault diagnosis method, which is shown in Figure 21.

Fault diagnosis method based on a two-stage RBF neural network

Full size image

Liu et al. [108, 109] improved the algorithm by using support vector machine regression (SVR) and the Elman neural network, which reduced the demand of the algorithm for data volume and avoided the risk of falling into a local minimum. Hu et al. [110, 111] and Cao et al. [112] also used the support vector machine (SVM) method to build the hydraulic system model. They also used the composite method to judge the deviation, which gets better diagnosis results in the case of small training samples.

From the methods listed above, it can be seen that the data-based fault diagnosis method has great flexibility because it uses a large amount of data to summarize the rules. By changing the method, the fault diagnosis can be achieved in different situations from a few sensors to a large number of sensors. However, because the electro-hydraulic valve and the hydraulic system are complex and nonlinear, they have a variety of fault forms and a large amount of data is required for model training. Because the hydraulic system is often used in the heavy load situation, lots of simulation works will consume too much energy and time, which increases the difficulty of using data-based fault diagnosis methods.

4.2.3

Model-Based Fault Diagnosis Methods

The model-based fault diagnosis method closely connects the fault characteristics with the model parameters; however, it is difficult to build an accurate mathematical model. Because the model has been validated by a lot of data, more accurate calculation results can be obtained. At the same time, the physical model is used to calculate the internal state of the system, which effectively reduces the requirements of the model correction data. Figure 22 provides an illustration of how it works.

The model-based fault diagnosis method

Full size image

In order to improve the accuracy of the model of the electro-hydraulic valve control cylinder system, Münchhof [89] used the least squares method. Even though the model is compensated in real-time, the fault diagnosis can be carried out by comparing the identified parameters with the parameters under normal working conditions [113]. Samadani et al. [114] also used the method of parameter estimation to identify the amplification coefficient of the electromagnetic force and the friction force of the valve spool movement in the electro-hydraulic servo valve. Another direct method is to directly use the open-loop model simulation to compare with the actual physical system signal and diagnose the fault through the obtained deviation [115,116,117,118]. Due to the lack of a mathematical mechanism to ensure the unbiasedness of the model, this method needs to use a more complex nonlinear model in the modeling process. In addition, more data is required to correct the model.

The parity space-based fault diagnosis method also uses the open-loop model to calculate the deviation signal. This method mainly uses historical data to establish the transfer equation. Although there is no acquisition signal to correct the model, the use of historical data improves the accuracy of the model. Münchhof [89] studied the fault diagnosis of the proportional servo control system with this method and the effectiveness of this method is verified for a variety of fault diagnoses.

In order to avoid the influence of the modeling-error for the open-loop model on fault diagnosis, most of the existing model-based fault diagnosis methods use the theoretical unbiased model as the contrast model to compensate for the model error. The common methods are the state observer and the state predictor.

In , Luenberger proposed an observer for linear systems, which became an important method for state estimation and fault diagnosis [119]. Min et al. [120] used a linear output observer to calculate the system deviation and diagnosed the abnormal response fault of the servo valve. Abou et al. [121] also used the linearization method to model the hydraulic system and set up the fuzzy fault diagnosis rules to diagnose the severity of the fault. Because a linear model produces the system error, the observer can design methods while considering the modeling error and the external interference, which has attracted the attention of researchers.

The first way to reduce the influence of the modeling error is to perfect the model and reduce the modeling error. Khan et al. [122] designed the feedback gain of an observer based on data and built a nonlinear observer to obtain the system deviation and diagnose the faults such as the change in the oil&#;s elastic modulus. Rezazadeh et al. [123] also used a similar method to diagnose the leakage in the electro-hydraulic proportional system.

The second way is to decouple the unknown factor and the known response law by feedback using the designed observer. Qiaoning Xu decoupled the interference term, which includes damping and the external load, along with the target input and the output feedback. This distinguishes the fault performance from the influence of unknown parameters [124, 125].

The third way is to use robust control algorithms, which uses feedback data to compensate for the model error and to improve the overall tracking performance of the model. Yao et al. [126, 127] proposed an adaptive robust observer design method based on the nonlinear coordinate transformation and an indirect adaptive strategy. Yao et al. [128] also proposed a state reconstruction method for nonlinear systems. Gayaka et al. [129] combined identification with the online diagnosis method as shown in Figure 23. In addition, Alwi et al. [130] and Liu [131] used a sliding mode observer to diagnose the valve failure in the electro-hydraulic servo system.

Fault diagnosis method based on the robust observer and the state reconstruction

Full size image

The method based on the state predictor mainly refers to the fault diagnosis method based on the recursive Bayesian filter algorithm, and the Kalman filter is a typical implementation form of the recursive Bayesian method. The basic structure of the Kalman filter is shown in Figure 24. Because the electro-hydraulic valve and its system are typical nonlinear systems, the application of a nonlinear Kalman filter for the fault diagnosis of electro-hydraulic directional valves has been introduced.

Structure diagram of the Kalman filter

Full size image

The state prediction of the nonlinear Kalman filter can be solved as the product of the nonlinear function and the Gaussian probability density function [132]; hence, the discrete Kalman filter method is mainly developed from two aspects: the approximation of the nonlinear function and the approximation of the Gaussian probability density function [133]. Wang et al. [134] used EKF (Extended Kalman Filter) to study the liquid elastic fault. Pulak Halder&#;s team used EKF with different acquisition signals for feedback during comprehensive fault diagnosis; this was mainly used for the diagnosis of sensor failure and leakage [135, 136]. An and Sepehri used EKF to segment the friction force, which has a good effect on the early diagnosis of leakage [137,138,139]. Due to the serious linearization error introduced by EKF, the flexibility of the application is limited.

The approach method to a Gaussian probability density function is a typical unscented Kalman filter method, which can improve the accuracy of the model [140, 141]. Sepasi et al. [142] used the mean absolute error of the pressure and displacement signals of the two chambers of the cylinder as the signal characteristics. This can obtain better fault diagnosis results for the valve port flow coefficient changes and other faults.

Nurmi and Mattila [143] studied the application of the UKF (Unscented Kalman Filter) algorithm in the fault diagnosis of the swing cylinder system, which is controlled by the four-position three-way proportional directional valve. The experimental results show that the proposed UKF fault diagnosis method can effectively diagnose the leakage even when the leakage flow is only 5% of the valve port flow. Nurmi and Mattila [144] extended the fault diagnosis method to the pilot proportional directional multi-valve cylinder control system with proportional reducing valves as the pilot stage in their later research. Experiments show that the UKF fault diagnosis method and adaptive threshold method can effectively avoid the impact of the impact signal on the fault diagnosis and achieve the detection of the small amplitude fault.

The communication interface that steps toward integration and compatibility makes the valve a component that has self-sensing, self-diagnosing, and IoT information interaction capabilities. Fault diagnosis is an important part of modern intelligent valves that are oriented to Industry 4.0. This can reduce the chance of an accident and is the basis for predicting the life of the valve. The combination of data-based and model-based methods will make the fault diagnosis process more accurate and efficient.

For more hydraulic flow control valveinformation, please contact us. We will provide professional answers.