The inertance hydraulic converter relies on fluid inertance to modulate flow or pressure and is considered to be a competitive alternative to the conventional proportional hydraulic system due to its potential advantage in efficiency. As the quantification of fluid inertance, the suction flow characteristic is the crucial performance indicator for efficiency improvement. To explore the discrepancy between the passive inertance hydraulic converter featured by the check valve and the active inertance hydraulic converter driven by an equivalent 2/3 way fast switching valve in regard to suction flow characteristics, analytical models of the inertance hydraulic converters were established in MATLAB/Simulink. The validated models of the respective suction components were incorporated in the overall analytical models and their suction flow characteristics were theoretically and experimentally discussed. The analytical predictions and experimental measurements for the current configurations indicated that the active inertance hydraulic converter yields a larger transient suction flow rate than that of the passive inertance hydraulic converter due to the difference of the respective suction components. The suction flow characteristic can be modulated using the supply pressure and duty cycle, which was confirmed by experimental measurements. In addition, the suction flow characteristics are heavily affected by the resistance of the suction flow passage and switching frequency. There is a compromise between the resistance and switching frequency for inertance hydraulic converters to achieve large suction flow rate.

To acquire a linear load pressure with the potential of energy-saving compared with conventional throttling hydraulic systems, a switched hydraulic circuit (referred to as a flow-dependent inertia hydraulic converter) featuring a flow-dependent inertial effect was proposed. In this device, the flowing fluid modulated by an equivalent fast switching valve allows the fluid inertance at the connection pipe and rotation inertia incorporated in the flywheel to be combined. The theoretical and simulation models of the flow-dependent inertia hydraulic converter were established, and then the flow-dependent inertia was discussed and quantified. The predictions of the load pressure, flywheel speed and system efficiency were confirmed by experimental measurements. Results indicated that the theoretical and simulation predictions match well with the experiments for this configuration. The pressure wave propagation effect within the connection pipe is responsible for the variation in the transient load pressure and flywheel speed that is initially modulated by the rotation inertia theoretically, which seriously questions the predecessor's hypothesis of using a compressible volume to model the upstream connection pipe. The parabolic mean suction flow induced by the flow-dependent inertia draw our attention to the available supply flowrate and duty cycle. Finally, the throttling loss and system efficiency that varies with the desired characteristics is superior to that of conventional throttling hydraulic systems theoretically, and an acceptable system efficiency was obtained under the premise of ignoring friction experimentally.

the aim of the overall research presented here was to investigate the motion of a wheeled mobile robot in an indoor (structured) setting while following a pre-set trajectory.An example of an application for this research would be automated maneuvering of a smart power wheelchair in a health care setting, such as a hospital, which would benefit the aging Canadian population.For the specific research reported here, the aim was to investigate the mobility of an assistive device for Sit-to-Stand (STS) operation and walking.A rehab robot was developed and was attached to a wheeled mobile robot to accomplish STS operation and to help walk a patient for rehabilitation.Four major phases of this research are: (i) design of the rehab robot, (ii) development of the control algorithm, (iii) experimentation, and (iv) navigation of the mobile robot and the rehab walker (robot).This research project can be extended to lower limb rehabilitation and design of a smart walker.It intensively studied current research and projects, designed a rehab robot and four control algorithms to help people in both STS and walking processes.Experiments were implemented, and the results indicated the effectiveness of the control algorithms and the prescribed navigation algorithm.According to the results, this project achieved the original goals to assist people in standing up and walking, and to navigate to a pre-set location.

In this paper we present an innovative space-variance approach named as "Space-Invariant Signature Algorithm (SISA)" for processing images from active systems, such as cancer cells, tumor growth, and dead cells, for the detection and localization of abnormalities at an early stage. In this paper, a SISA processing algorithm is developed, and this algorithm is tested on animal tissues such as pigs and chicken tissues. The abnormality in an active system can be defined as the obstacle or a failure which impedes the activities in tissues such as smooth flow of blood or electrical signals etc. Due to this impeding nature of the abnormality, some parameter perturbations are induced. In this paper using the SISA approach, these perturbations were detected in a preliminary experiment on animal tissues. The degree and position of the space-variance helps us in the detection and localization of abnormality even at an early (incipient) stage. The space-variance signature pattern is named as a SISA signature pattern'. In the absence of any abnormality, the signature pattern is space invariant, whereas, in the presences of any abnormality, the SISA signature pattern varies in the space (space-variant). The basic experimental studies on animal tissues using ultra sound imaging strongly suggest a possible use of the SISA approach as a non-invasive method for the detection and localization of abnormalities in biological tissues such as cancer cells non invasively.

In this paper we present an innovative space-variance approach named as “Space-Invariant Signature Algorithm (SISA)” for processing images from active systems, such as the liquid vibrating active system, cancer cells, tumor growth, and dead cells, for the detection and localization of abnormalities at an incipient stage. In this paper, a SISA processing algorithm is developed, and this algorithm is tested on a liquid vibrating active system. The abnormality in an active system can be defined as the obstacle or a failure, which impedes the activities such as vibrations, smooth flow of blood or electrical signals etc. Due to this impeding nature of the abnormality, some parameter perturbations are induced. In this paper using the SISA approach, these perturbations were detected in a preliminary experiment on a liquid vibrating active system. The degree and position of the space-variance helps us in the detection and localization of abnormality even at an early (incipient) stage. The space-variance signature pattern is named as `SISA signature pattern'. In the absence of any abnormality, the signature pattern is spaceinvariant, whereas, in the presence of any abnormality, the SISA signature pattern varies in the space (space-variant). The basic experimental studies on a liquid vibrating active system strongly suggest a possible use of the SISA approach as a non-invasive method for the detection and localization of abnormalities in biological tissues such as cancer cells non-invasively, and this work will be reported in another paper.

It is well known that the presence of entrained air bubbles in hydraulic oil can significantly reduce the effective bulk modulus of hydraulic oil. The effective bulk modulus of a mixture of oil and air as pressure changes is considerably different than when the oil and air are not mixed. Theoretical models have been proposed in the literature to simulate the pressure sensitivity of the effective bulk modulus of this mixture. However, limited amounts of experimental data are available to prove the validity of the models under various operating conditions. The major factors that affect pressure sensitivity of the effective bulk modulus of the mixture are the amount of air bubbles, their size and the distribution, and rate of compression of the mixture. An experimental apparatus was designed to investigate the effect of these variables on the effective bulk modulus of the mixture. The experimental results were compared with existing theoretical models, and it was found that the theoretical models only matched the experimental data under specific conditions. The purpose of this paper is to specify the conditions in which the current theoretical models can be used to represent the real behavior of the pressure sensitivity of the effective bulk modulus of the mixture. Additionally, a new theoretical model is proposed for situations where the current models fail to truly represent the experimental data.

T-joint connections are used extensively in industry as parts of machine components and structures. The T-joint connection is typically constructed through the welding of its tubular members, with significant stress and strain concentrations occurring at the toe of the weld under loadings. In this paper, a welded T-joint connection of square hollow-section (SHS) tubes subjected to a multi-axial state of stress is examined both numerically and experimentally. The hot spot strains and stresses in the connection are determined through a detailed finite element (FE) analysis of the joint. The weld geometry is accurately modelled using FE. To model the weld, several full-scale welded T-joints were cut at the connection to obtain the size and depth of penetration of the weld.For the experimental study, a test rig with a hydraulic actuator capable of applying both static and cyclic loadings is designed and used. Strain gauges are installed at several locations on the joint to validate the FE model. The verified FE model is then used to study the through-the-thickness stress distributions of the tubes. It is shown that the membrane stresses which occur at the mid-surface of the tubes remain similar regardless of the weld geometry. The weld geometry only affects the bending stresses. It is also shown that the stress concentrations are highly localized at the vicinity of the weld toe. At a distance of about half of the weld thickness from the weld toe, the effect of the weld geometry on the bending stresses becomes insignificant as well. To reduce the stress concentrations at the T-joint, plate reinforcements are used in a number of different arrangements and dimensions to increase the load-carrying capacity of the connection.

Modeling of fluid power components and systems is a challenge at the best of times. As knowledge of physical components becomes more focused on micro-properties, the describing equations become more comprehensive and reflect dynamic performance that in the past has not been possible to accurately model. Usually the equations are very nonlinear which make the resulting describing differential equations difficult to solve. But perhaps a more substantial problem in modeling is being able to assign values to a large number of parameters which are used to describe the system. A further complication arises in that the parameters are functions of operating conditions. An alternate approach to modeling is to use a "black box" in which equations which reflect physical properties are replaced by input/output data relationships. Neural networks are common forms of the mechanism which govern the foundation of the black box in which neurons are trained to "learn" the input/output relationships of a physical system through training and testing processes. The authors have examined various types of neural network morphologies in an attempt to model a load sensing pump using this approach. In a paper submitted at the Bath workshop last year, some success was reported in capturing the dynamics of an experiential load sensing pump but it was concluded that a different network morphology was necessary to make the approach practical and more efficient. In this paper the authors introduce a recursive generalized neural network (RGNN) form that appears to overcome some of the difficulties that were encountered in earlier studies. In addition, the network was trained using a non gradient technique (the complex method) which runs counter to the gradient-based foundation of most back-propagation techniques. Experimental data used in these previous studies were used to train the recursive generalized neural network and under testing, accurately reflected both the static and the dynamic characteristics of the pump. The key to the successful training of the network was the setting up of experimental test procedures which would capture the dynamic performance of the unit at expected operating conditions. The result of this study is significant because the designed neural network is very efficient and lends itself to integration to other conventional simulation packages.

The objective of this work was to design and implement an experimental hydraulic system that simulates joint flexibility with the ability of changing the joint flexibility's parameters. Such a system could facilitate future control studies by reducing investigation time and implementation cost of research. It could also be used to test the performance of different strategies to control the movement of flexible joint manipulators. A hydraulic rotary servo actuator was used to simulate the action of a flexible joint robot manipulator. A challenging task since the control of angular acceleration was required. A single-rigid-link, elastic-joint robot manipulator was modeled using Matlab®. Joint flexibility parameters such as stiffness and damping, could be easily changed. This simulation could be referred to as "function generator" to drive the hydraulic flexible joint robot. In this study the angular acceleration was used as the input to the hydraulic rotary actuator and the objective was to make the hydraulic system follow the desired acceleration in the frequency range specified. The hydraulic system consisted of a servo valve and rotary actuator. A hydraulic actuator robot was built and tested. The results indicate that if the input signal had a frequency in the range of 5 to 15 Hz and damping ratio of 0.1, the experimental setup was able to reproduce the input signal with acceptable accuracy. Because of the inherent noise associated with the measurement of acceleration and some severe nonlinearities in the rotary actuator, control of the experimental test system using classical methods was not as successful as had been anticipated. This was a first stage in a series of studies and the results provide insight for the future application of more sophisticated control schemes.

A load sensing (LS) system is one in which the pump flow is regulated to keep the pressure drop across an orifice constant and independent of any variation in the load pressure. This ensures that the pressure loss across the orifice is kept to a minimum, thereby increasing efficiency. An LS regulator spool is used to sense the pressure drop across the orifice to control pump delivery. The spool can be underlapped, critically lapped or overlapped. As a trade-off between efficiency and dynamic response, the LS spool is usually critically lapped. This results in a nonlinear model that is sensitive to operating regions.In this paper, a review of published literature on LS systems is briefly summarized. An LS system model is developed and linearized. Procedures to solve these very complex equations are introduced. Because load sensing systems require pressure feedback, stability can often be an issue. Analysis of these systems to determine the steady state and dynamic performance is very difficult to do because of the dependency of the models on the operating point. Linearized models which reflect a methodology to account for changing operating conditions have been developed and have established three distinct regions of operation (labeled "Conditions I, II, and III"). This paper presents the experimental nature of these conditions and provides experimental evidence that the models so derived are valid over certain frequency ranges. The objective of this paper, then, was to establish confidence in the models by examining frequency response performance under these three distinct conditions. The results show that good agreement does exist between the models and their physical counterparts and establishes limitations thereof.This research can assist in the design or optimization of an LS system and help in the development of advanced control strategies for obtaining further efficiency within certain dynamic performance constraints.

A pressure-compensated valve (PC valve) is a type of flow control device that is a combination of a control orifice and a compensator (often called a hydrostat). The compensator orifice modulates its opening to maintain a constant pressure drop across the control orifice. In other words, the PC valve is so designed that the flow rate through the valve is governed only by the opening of the control orifice and is independent of the total pressure drop across the valve. Because of the high nonlinearities associated with this type of valve, it is impossible, in practice, to design such a valve where the flow rate is completely unaffected by the pressure drop across the valve. In this paper the effect of the nonlinearities on the performance of the PC valve is investigated. First, a generic nonlinear model of a PC valve is developed. Using this model, all possible operating conditions can be determined. Then a linearized model is developed and used to analyze the dynamic behavior of the PC valve. The model can then be used to evaluate and improve the design and operation of the valve for specific applications.

A pressure compensated valve (PC valve) is a type of flow control device that is a combination of a control orifice and a compensator (often called a hydrostat). The compensator orifice modulates its opening to maintain a constant pressure drop across the control orifice. In other words, the PC valve is so designed that the flow rate through the valve is governed only by the opening of the control orifice and is independent of the total pressure drop across the valve. Because of the high non-linearities associated with this type of valve, it is impossible, in practice, to design such a valve where the flow rate is completely unaffected by the pressure drop across the valve. In this paper, the effect of the non-linerities on the performance of the PC valve is investigated. First, a generic non-liner model of a PC valve is developed. Using this model, all possible operating conditions can be determined. Then a linearized model is developed and used to analyze the dynamic behavior of the PC valve. The model can then be used to optimize the design and operation of the valve for specific applications.

The ability to monitor the progress of a fault in its early stages in an electro-hydraulic valve is desirable because it may be possible to project when a fault has progressed to the point where it should be replaced before a failure actually occurs. Knowledge of existing or non-existing faults can be utilized directly in a maintenance program; unnecessary shut-downs can be avoided, fault diagnosis can be simplified, component life can be extended, etc. This paper outlines a fault detection procedure for single-stage proportional valves. The procedure employs the estimation and tracking of states that effectively describe the overall health of the valve. A technique is developed for monitoring any statistically significant change in these states during the life of the valve. The valve spool is forced to follow a unique waveform during scheduled intervals at which coil current and voltage are simultaneously recorded. A cost-effective hydraulic circuit is introduced that isolates the valve from the load, facilitating safe acquisition of pertinent sensor data. A practical procedure for acquiring data is introduced which is invisible to the operator and guarantees repeatability of environmental conditions. Background processing utilizes neural network theory to segregate measured current into three states representative of coil reluctance, coil hysteresis and spool friction. Data acquisition is repeated 10 times evaluating a mean of each state from the resulting population. Another mean of each state is evaluated from an additional 10 sets of data acquired at an identical temperature. Statistical comparison of the means proves repeatability and hence preliminary feasibility. Comparison to another set of means evaluated at a different temperature showed unacceptable variance. The importance of acquiring data at a consistent temperature is therefore verified.

Abstract Modelling hydraulic control systems that contain flow modulation valves is highly influenced by the accuracy of the equation describing flow through an orifice. Classically, the basic orifice flow equation is expressed as the product of cross-sectional area, the square root of the pressure drop across the orifice and a “flow discharge coefficient”, which is often assumed constant. However, at small Reynolds numbers (such the case of valve pilot stage orifices), the discharge coefficient of the flow equation is not constant. Further, the relationship between the flow cross-sectional area and the orifice opening are extremely complex due to clearances, chamfers, and other factors as a result of machining limitations. In this work, a novel modification to the flow cross-sectional area is introduced and the resulting closed form of the flow equation is presented. As a secondary benefit, an analytical form of the orifice flow gain and flow pressure coefficient can be obtained. This closed form equation greatly facilitates the transient and steady state analysis of low flow regions at small or null point operating regions of spool valve.

A load sensing system is one in which the pump flow is adjusted to keep pressure across an orifice constant and independent of any variation in the load pressure. This ensures that the pressure losses across the orifice are kept to a minimum which increases efficiency substantially. Because the system is closed loop, stability can become a problem. To establish stability bounds, linearized analysis is often employed. However, to do this, operating points of all linearized parameters and coefficients must be established as a function of certain parameters such as load pressure. This can only be done by solving a series of nonlinear algebraic equations. This paper presents a set of equations for three special conditions. The experimental verification of operating points that are predicted for such a load sensing system is presented. The three regions are established theoretically and are verified experimentally. It is found that the operating points undergo a noticeable change when in transition from one region to another (as dictated by variations in load pressure or orifice area). It was also found that the agreement between the predicted and measured operating points was quite satisfactory and could be used with confidence in future studies.

The electrical-mechanical interfaces of servo valves and proportional valves display electromagnetic nonlinearities, such as hysteresis and saturation. In many cases, these nonlinearities have usually unfavorable effects on the performance of the whole control system. For this reason the concept of stage proportional control is put forward in this paper. The fundamental idea of the stage proportional control is to divide the whole stroke of the hydraulic (pneumatic) control valve into a number of stages in which independent proportional control is carried out.In this way, the electromagnetic nonlinearities can be greatly reduced. The stage proportional control can be realized by using a linear or rotary stepper motor controlled in a continuous mode. Experimental results demonstrated the electromagnetic nonlinearities associated with proportional control are indeed reduced and the position stiness of the spool and the resolution accuracy are also enhanced. An application example of the stage proportional control is also presented in this paper.

Many control schemes, simple or sophisticated, utilize high performance electro-hydraulic components as an interface to mechanical hardware. Failure of an electro-hydraulic component in these applications may impact safety, maintenance schedules and/or productivity of the overall operation. A condition monitoring scheme that would measure the performance or health of critical electro-hydraulic components would be desirable in addressing the above mentioned concerns.Present research at the University of Saskatchewan involves the feasibility of acquiring data from a proportional solenoid valve on-line for the purpose of condition monitoring. Once data have been acquired (specifically spool displacement and spool differential pressure), an algorithm can be implemented to estimate the desired valve parameters. Some of the parameters that affect the performance of the main stage valve spool which may be used in a condition monitoring scheme include area orifice gradient, spring rates, spring pretensions and friction.This paper considers the use of individual 'neuron-like' structures to estimate each valve parameter. Each neuron trains to a specific valve parameter with its output being the estimated parameter. The individual neurons act and train independently; however, their outputs are integrated to compute the main spool differential pressure. This differential pressure calculation is compared to the measured value and the difference is used to train each neural structure. Experimental results are presented which demonstrate the accuracy and feasibility of the procedure.

In many advanced control schemes, it is necessary to have a mathematical model of the valve in order to define an appropriate control algorithm. Central to the modeling process is the measurement or assignment of the valve's parameter or coefficient values. The manufacturer can often specify some of the parameter values such as spring constants, spool mass, spool areas, etc. However, friction, flow reaction force, and spool metering area values, to name a few, are very difficult to define or measure and must be approximated in some fashion. This has been the subject of much interest in the past. These estimation approaches have tremendous potential in that the various coefficient and parameter values could also be used to monitor the health or state of the valve over time, a process commonly referred to as condition monitoring or fault diagnosis.A problem common to parameter estimation procedures when used in condition monitoring/fault diagnosis schemes is to be able to duplicate the operating conditions when the "norm" and on-line estimations are made. This paper considers a possible approach to establishing this norm. In addition, a simple method of estimating parameters of the main stage of a proportional valve is also considered. Experimental results are presented which demonstrate the reliability of the method, and the approach for actual implementation is discussed.

The use of dynamic neural networks to model and control dynamic systems is of great importance in the control paradigm. The intent of this paper is to use one such dynamic neural structure, namely the recurrent neural network, to drive unknown nonlinear systems to follow the desired trajectories. The learning scheme employed for this task consists of a conventional proportional-plus-derivative (PD) controller in the feedback loop and the recurrent neural network in the feedforward path. Once the convergence is achieved, the recurrent neural network approximates the inverse-dynamics model of the plant under control. The PD controller, on the other hand, guarantees the stability of the learning scheme. The effectiveness of this learning scheme is demonstrated through computer simulations and an experimental setup that demonstrates the balancing of a two-wheeled robot