This study presents models of an electro-hydraulic valve derived from physical principles and neural network techniques. Input-output models are constructed using experimental data from a hydraulic press machine in a steel manufacturing plant and as a plant of a closed loop system. The models are candidates for digital twins in the steel manufacturing plant. The physical model is derived based on fluid mechanics, fluid dynamics, and electronic principles. Two LSTM neural network models denoted NN1 and NN2 are employed for modeling the valve. The parameter estimation for each neural network model is conducted using distinct training datasets. This work compares the validation results of the models in the time and frequency domain. Both the physical model and NNs capture the main behavior of the valve. However, NNs have lower mean square error (MSE) compared to the physical model. NN2, trained the model using different operating conditions, is capable of modeling non-linearities (seen at high frequencies) and leakage effects of the system, that are not captured by NN1 and physical models. The fault caused by leakage is seen in the MSE vs cycles for the physical and NN1 models.

Determining moisture content (MC) in wood chips finds its application in many industries including energy production. In this letter, we aim to develop an automated method for determining MC in woodchips using ultra-wideband (UWB) radio signals and machine learning algorithms. Firstly, to acquire UWB signals through wood chips on conveyor belts in industrial plants, we use measurement devices with a radio transmitter and receiver, and a laser sensor to determine the thickness of the wood chips. UWB and laser data corresponding to 1923 samples from four power plants is acquired. Secondly, we extract the amplitude and delay-based features, and these are finally, fed to three different machine learning algorithms namely; linear regression, artificial neural network (ANN), and ensemble trees to determine the MC. The proposed method achieves best results when the ANN is used. More specifically, our method achieves a mean absolute error (MAE) of 2.75% when the features from both UWB and laser sensors are used for determining MC. MAE of 3.95% is achieved when features only from UWB data (without the laser) are used for determining MC. Our results for industrial data suggest that the proposed method is effective for determining MC in industrial applications.

Digital twins are an emerging technology that can be harnessed for the digitalization of the industry. Steel industry systems contain a large number of electro-hydraulic components as proportional valves. An input–output model for a water proportional cartridge valve was derived from physical modeling based on fluid mechanics, dynamics, and electrical principles. The valve is a two-stage valve with two two/two-way water proportional valves as the pilot stage and a marginally stable poppet-type cartridge valve as the main valve. To our knowledge, this is the first time that an input–output model was derived for a two-stage proportional cartridge valve with a marginally stable main valve. The orifice equation, which is based on Bernoulli principles, was approximated by a polynomial, which made the parameter estimation easier and modeling possible without measuring the pressure of the varying control volume, in contrast with previous studies of similar types of valves situated in the pilot stage part of the valve. This work complements previous studies of similar types of valves in two ways: (1) data were collected when the valve was operating in a closed loop and (2) data were collected when the valve was part of a press mill machine in a steel manufacturing plant. Model parameters were identified from data from these operating conditions. The parameters of the input–output model were estimated by convex optimization with physical constraints to overcome the problems caused by poor system excitation. For comparison, a simple linear model was derived and the least squares method was used for the parameter estimation. A thorough estimation of the parameters’ relative errors is presented. The model contains five parameters related to the design parameters of the valve. The modeled position output was in good agreement with experimental data for the training and test data. The model can be used for the real-time monitoring of the valve’s status by the model parameters. One of the model parameters varied linearly with the production cycles. Thus, the aging of the valve can be monitored.

We present a printed lens for radar applications. The structure of the presented lens consists of an array of modified micro-strip lines, which is positioned in the antenna’s aperture on the same planar substrate. Simulations show that the gain and directivity increase with the proposed lens in a wide band frequency band. The proposed design is insensitive to rotation of the antenna. This paper focuses on real industrial applications and problems. Further, we show that the lens can be used to improve the object detection ability of an ultrawide band radar system, which is used in industrial applications such as non-destructive monitoring of built-structures and for use in the renovation process. The signal to noise ratio is improved. Furthermore, we show how the microwave lens can also be used to reduce the clutter in applications where the complex refractive index of objects is determined. Further, different simulated results (for different cases) are compared, presented and concluded.

A digital twin of a mechanical system (a pair of axial rolls in a ring mill used in a steel plant) with poles close to the unit circle and the real axis in the discrete pole-zero map was built. The system was excited by a signal concentrated in the low-frequency band. For this particular case, it is shown that the ad-hoc combination of ARMAX and orthonormal basis filter model structures outperform model structures based on either ARMAX or orthonormal basis functions when estimating the poles of the basis by analyzing the data in the frequency domain. The followed modelling methodology of the system is described in detail to help replicate the work for similar systems in the steel industry. Real production data from a steel plant were used in contrast to previous studies, where the combination of ARX and ARMAX with orthonormal basis filter model structures was evaluated using simulated data instead of real data. We believe that the resultant model can be used when having systems with poles close to the unit circle and real axis and poor excited input signal concentrated in the low frequency band. The resultant model can be used for condition monitoring and failure detection.

A measurement system for the measurement of microwave dielectric properties of mineral compounds at temperatures up to +1000°C is presented. It includes the simultaneous measurement of mass and temperature. Samples volumes in the range 0.01 to 0.1 m 3 can be studied. The system comprises a heat reaction chamber on a mass scale with mounted ultra-wideband (UWB) radio sensors and temperature probes. The complex refractive index is determined from the UWB signals using a technique with windowing to suppress interference and fitting of a modelled signal to the experimental ones. The developed method is validated by measuring the complex refractive index of water from +82°C down to +23°C and comparing with literature values. The systems is used to study calcination of limestone, i.e. the chemical decomposition of CaCO 3 to CaO and CO 2 when heated up to +1000°C. The chemical decomposition is clearly seen as a decrease in mass and as significant changes in the complex refractive index. The system could be used also for other mineral compounds and other types of materials.

In the present work, a digital twin of a radial-axial ring rolling machine was built by modelling the time series of the positions of the tools and control signals rather than the metrics of the produced rings, as performed in previous studies. Real data from the industry was used for modelling. The used model selection methodology is shown in detail to replicate such work for similar systems in the steel industry. The modelling results of ARX, ARMAX and orthonormal basis model structures are shown; additionally, they were validated considering SISO and MIMO systems. The modelling results were better when the subsystems considered were ARMAX and MISO than when ARX and SISO were taken into consideration. The best modelling results were obtained when physical knowledge was included in the model structure. Lastly, it was found that the model error of the horizontal subsystem could be used for predictive maintenance.

An ultra-wide band measurement method for determining the complex refractive index of large-volume objects is presented. The method is intended for industrial non-destructive testing. It uses a frequency-domain technique in which transmitted radio pulses are analyzed and the effects of near field and coupling on the determined refractive index are compensated. Measurements were performed in an industrial setup with electromagnetic sensors buried in the object. The results are presented for woodchips as an object. The refractive index was experimentally determined in the frequency range 0.5-3.0 GHz. Additionally, we designed and manufactured planar quasi-differential elliptical-antennas as electromagnetic sensors. The results from the industrial measurement setup were compared with the results of the laboratory setup, in which the sensors were placed outside the test box and near field and coupling effects could be neglected. The complex refractive index determined for the two setups was in good agreement, which corroborates the proposed method for compensating for coupling and near-field effects. The complex refractive index of woodchips changes with the moisture content. It is experimental verified using the industrial setup that the moisture content can be determined with a 2 percent error.

The complex microwave permittivity-including anisotropy- of wood chips of softwood has been measured for different moisture contents in the band 0.75 to 2.5 GHz using an ultra-wide band radio transmission technique. The real and imaginary parts increase monotonically with moisture content. The wood chips are oriented by gravity, which gives anisotropic permittivity. The anisotropy ratio of the real part increases from 1.1 to 1.6 with moisture content from 0 to 120%. The anisotropy ratio of the imaginary part is around 2.5 at all moisture contents. Effective medium models were used to model the permittivity. The Bruggeman, and two versions of the Maxwell Garnett model gave good results at low moisture content (below the fiber saturation point). Above the fiber saturation point only the Bruggeman model gave results in agreement with experiments. The difference in model performance suggests that the free water does not follow the wood chips geometry.

Experimental characterization of an amplifier's nonlinear properties at 24 GHz is presented in single and dual-band operation using orthogonal frequency-division multiplex signals. A test system for characterizing an amplifier's nonlinear properties at millimeter-wave frequencies for single and dual-band excitation is presented. The use of standard instrument enables a feasible test system. Analytical expressions based on a statistical analysis of signals and hardware impairments were used to analyze the experimental data versus power level and found to describe well the experimental results, including inter- and cross-modulation distortion. Parameters are derived that could be used in system studies.