Mini-batch optimization enables training of ODE models on large-scale datasets

Quantitative dynamic models are widely used to study cellular signal processing. A critical step in modelling is the estimation of unknown model parameters from experimental data. As model sizes and datasets are steadily growing, established parameter optimization approaches for mechanistic models become computationally extremely challenging. Mini-batch optimization methods, as employed in deep learning, have better scaling properties. In this work, we adapt, apply, and benchmark mini-batch optimization for ordinary differential equation (ODE) models, thereby establishing a direct link between dynamic modelling and machine learning. On our main application example, a large-scale model of cancer signaling, we benchmark mini-batch optimization against established methods, achieving better optimization results and reducing computation by more than an order of magnitude. We expect that our work will serve as a first step towards mini-batch optimization tailored to ODE models and enable modelling of even larger and more complex systems than what is currently possible.

The effect of sea surface temperature on the structure and connectivity of species landings interaction networks in a multispecies recreational fishery.

Multispecies fisheries, particularly those that routinely adapt the timing, location, and methods of fishing to prioritize fishery targets, present a challenge to traditional single-species management approaches. Efforts to develop robust management for multispecies fisheries require an understanding of how priorities drive the network of interactions between catch of different species, especially given the added challenges presented by climate change. Using 35 years of landings data from a southern California recreational fishery, we leveraged empirical dynamic modelling methods to construct causal interaction networks among the main species targeted by the fishery. We found strong evidence for dependencies among species landings time series driven by apparent hierarchical catch preference within the fishery. In addition, by parsing the landings time series into anomalously cool, normal, and anomalously warm regimes (the last reflecting ocean temperatures anticipated by 2040), we found that network complexity was highest during warm periods. Our findings suggest that as ocean temperatures continue to rise, so too will the risk of unintended consequences from single species management in this multispecies fishery.

The Case Experience of Integrating the SDGs into Corporate Strategies for Financial Risk Management Based on Social Responsibility (with the Example of Russian TNCs)

The motivation of this research consists in the following: the traditional commercial approach to financial risk management amid economic crises implies the reduction of corporate social responsibility, based on the assumption that this responsibility raises the financial risk of business. Due to this, the contribution of business to the achievement of the SDGs is not stable and is often negative, since practices of business management contradict the SDGs in crisis periods and hinder their achievement in society and the economy. However, the refusal from corporate social responsibility during a crisis does not guarantee the following increase in the level of business development in the period of stability. A study of the case experience of integrating the SDGs into corporate strategies of the largest Russian companies during the COVID-19 crisis improved the understanding of the contribution of corporate social responsibility to financial risk management of the business. Dynamic modelling showed that, in a crisis period, corporate social responsibility leads to a reduction of the financial risks of business—it is commercially profitable, similarly to the phase of stability, and critically important. Based on this, an alternative (new) approach to financial risk management is developed, which allows raising the effectiveness of this management amid economic crises (including the COVID-19 crisis) through the integration of the SDGs into corporate strategies and the manifestation of high social responsibility during crises.

Dynamic Modelling of Piston the Motion in Combustion Engines

The presented work discusses a methodology for analysis of noise emissions from a diesel engine. A numerical model of the piston motion, analyzing its lateral, reciprocating and rotation motion, has been presented in order to investigate the lateral motion of the piston skirt assembly and resulting vibrations induced as a result of these motions in the engine block. Various parameters of modal analysis were obtained using the mobility analysis. The presented methodology was validated by data obtained from a diesel engine test set up. The predicted results matched well with those of measured data, hence validating the presented scheme.

Experimental validation of a dynamic modelling of a Reversible Solid Oxide Cells (rSOCs)

The integration of Hydrogen technologies in different end-uses such as transport, electric microgrids, residential and industrial applications, will increase exponentially soon. Hydrogen as energy carrier allows more favourable energy conversion than other conventional systems and is crucial in worldwide decarbonize end uses. The production of green hydrogen, using RES, is a key area for the evolution of this technology. In this context, SWITCH is a Horizon 2020 European Project that aims to design, build and test an in-situ fully integrated and continuous multisource hydrogen production system, based on solid oxide cell technology. Reversible Solid Oxide Cell (rSOCs) technologies allow to convert renewable energy as hydrogen in the power-to-gas application (P2G) and in reversible mode is able to produce electricity from hydrogen stored, power-to-power application (P2P). rSOCs are really interesting to stabilize the random nature of RES because a combined electrolysis and fuel cell system should be able to switch between the two modes as quickly as possible in order to optimize the integration and the use of RES. However, rSOCs need a complex BoP from the thermal point of view, able to guarantee high efficiency even at partial load mode as well as easy start-up and shutdown procedures. In this work, a Stack Box Module dynamic model was developed in Modelica environment as a dynamic tool for the definition and optimization of BoP requirements. Stack model was validated in SOFC (Solid Oxide Fuel Cell) and SOE (Solid Oxide Electrolyser). The results of the simulation provide verification of the technical/thermodynamic behaviour and flexibility of a stack box of 70 cells. Dynamic modelling allows to evaluate the effect of the reagent inlet temperatures on the operation and hydrogen production/consumption in terms of yield as well as the transients between the different operative modes. Model has been validated by experimental measurements performed in the laboratory. In particular, the kinetics of the reactions governing steam methane reforming (SMR) was considered from data found in the literature, while the ASR (Area Specific Resistance) value was calibrated according to experimental data. The results of the dynamic model show as model can be a useful design and optimization tool for the SOCs technology.

Effect of novel continuous friction model on nonlinear dynamics of the mechanisms with clearance joint

A general methodology for the dynamic modelling and analysis of planar multi-body systems with a continuous friction model in joint clearance is presented. Joint clearance is the critical factor that influences the dynamic response and the performance of mechanisms for high-speed application. In light of recent developments in the joint clearance studies, the number of contact force models has been introduced with ignoring friction continuity. The selection of an appropriate continuous friction model is still challenging and essential, which requires further development. Therefore, a perfect continuous friction model, including the Stribeck effect, static, dynamic and viscous friction terms, is proposed and validated. Investigating the dynamic modelling and analysis of double rocker four-bar linkage mechanisms with frictional revolute clearance joints is presented to investigate friction models' effect when surfaces collide with a non-zero tangential velocity. Unlike the smooth crank input mechanism, a double rocker four-bar linkage mechanism is analysed as a challenging problem in the impact mode. Resolving this concern, the novel friction model avoids discontinuity at zero velocity considering the accurate static friction zone. The results reveal that the novel friction model, compared with the piecewise friction model, is more effective in reflecting the mechanical systems' dynamic behaviour. In order to grasp the nonlinear characteristics of the high-speed four-bar linkage mechanism with our model in joint clearance, the Poincaré portrait, and Fast Fourier transformation plot are employed. It is proved that chaos exists in the dynamic response with the influence of the restitution coefficients and kinetic coefficient of friction.

Dynamic Modelling and Advanced Process Control of Power Block for a Parabolic Trough Solar Power Plant

A fundamental task in the dynamic simulation of parabolic trough power plants (PTPP) is to understand the behavior of the system physics and control loops in the presence of weather variations. This study provides a detailed description of the advanced controllers used in the power block (PB) of a 50 MWel parabolic trough power plant (PTPP). The PB model is achieved using APROS software based on the actual specifications of the existing power plant. To verify the behaviour of the PB model, a comparison between the simulated results and given real data is documented depending on a previous study, and the results indicate a reasonable degree of correspondence. The purpose of this study is to create reference models for the PB. Thereby, developers and engineers will have a better understanding of the state of the art of advanced control loops in these power plants. Moreover, these types of models can be used to specify the most suitable mode of operation for the power plant. In addition, this study gives an overview of dynamic simulation for the design, optimisation and development of power blocks in parabolic trough power plants.