Comparative study of neural network based and white box model predictive control for a room temperature control application

On a global scale, buildings are a major cause for primary energy consumption. Since buildings are complex multiple input multiple output systems and characterized by slow dynamics, model predictive control is a promising approach to reduce building energy consumption. Due to the high individual modeling effort model predictive control lacks practical applicability. For that reason black box process models are gaining more and more interest in scientific literature. In this work we evaluate the performance of an ANN based controller against a white box controller with perfect knowledge. We show that the data driven controller achieves a similar control quality as the white box controller. We initially train the data driven controller in 20 days and then employ an online learning strategy to continuously improve the control quality.

Study on Pollution Cost Control Model under Asymmetric Information Based on Principal Agent

Pollution cost control is key to solve pollution problem. The paper takes pollution control cost of pollution control contract between management authority and pollutant discharge enterprise as research object, considers pollution control quality level, pollution control quality inspection and pollution control cost model, and establishes pollution control cost model of management authority and pollutant discharge enterprise, including rational constraints of pollutant discharge enterprise. And it analyzes principal-agent relationship between the two under condition of asymmetric information, and un-observability of pollution control level is shown as hiding information of sewage enterprises. In essence, it is problem of adverse selection in principal-agent. Pollution control cost of management is objective function. The first order condition of pollution control cost of sewage enterprise is transformed into state space equation, and optimal control of problem is solved by using maximum principle. In particular, management authority, as principal, uses pollution control provisions to reward, punish and encourage pollutant discharge enterprises as agents.

Brain perfusion in type 1 diabetes and cognitive dysfunction

The purpose. To evaluate brain microcirculation using contrast and non-contrast magnetic resonance perfusion in patients with type 1 diabetes and cognitive dysfunction.Material and methods. The study complies with generally accepted ethical rules. The study included 45 patients with type 1 diabetes and cognitive dysfunction and 20 patients without. Every patient included in the study was continuously monitoring glycemia with evaluation of variability coefficients. MRI was performed using a magnetic resonance imaging scanner Signa Creator “E”, GE Healthcare, 1.5 Tl, China: methods – dynamic contrast (“Gadovist”, w/w, bolusno, 5 ml) and arterial spin marks. SPSS Statistic software package was used for statistical analysis.Results. Blood flow decreased in patients with type 1 type and cognitive dysfunction in the areas of white and gray matter of frontal, occipital and temporal lobes, and shells (p ≤ 0.05). According to perfusion data, hyperglycemia and the following glycemic variability indices have the greatest influence on cortical structures: glycemic index of prolonged glycemia increase, risk of hyperglycemia and hypoglycemia, glycemic rate of change, glycemic control quality indicator, and in case of non-contrast glycemic control quality and glycemic rate of change, risks of hypo and hyperglycemia. The main factors of changes in brain microcirculation are episodes of severe hypoglycemia in the anamnesis, duration of the disease, arterial hypertension, high cholesterol levels. No separate markers for the evaluation of cognitive disturbances in type 1 diabetes were revealed.Conclusions. The basis of microcirculatory brain disorders in type 1 diabetes is the level of HbA1c and variability of glycemia as well as acute complications, duration of diabetes and associated conditions (arterial hypertension and hypercholesterolemia). The most important data were obtained during contrast perfusion.

Media coverage and financial restatements: evidence from China

PurposeDrawing on the premise that the media play a vital corporate governance role, this paper aims to investigate the association between media coverage and financial report restatements.Design/methodology/approachBased on a sample of Chinese listed companies over the period 2011–2015, the authors use ordinary least squares regression as well as a number of additional tests. To mitigate the endogeneity issue, the authors use a two-stage Heckman test and a propensity score matching model.FindingsThe authors document a negative and significant association between media coverage and restatements, suggesting that firms with high media coverage engage less in financial restatements. The authors further explore the moderating effects of internal control quality and state ownership on the association between media coverage and restatements. Regression results reveal that the governance role of the media is more pronounced for state-owned enterprises than for private firms. However, no significant difference in the disciplining effect of media coverage is found for firms with high, versus low, internal control quality.Originality/valueThe role of the media in corporate governance and financial reporting quality has been well documented. In emerging economies, such a role has been overlooked. As a result, the purpose of this study is to fill that void. Furthermore, prior research ignores the impacts of state ownership and the internal control environment on the media's governance role.

FINANCING CONSTRAINTS, INTERNAL CONTROL QUALITY AND COST STICKINESS

Managers think that retaining resources is more effective than rebuilding resources after exhausting them. However, financing constraints have brought great uncertainty to this resource decision-making implemented by managers. Data of manufacturing listed firms in China from 2009 to 2017 are used here to explore the impact of financing constraints on cost stickiness. This paper finds that internal financing constraints have a significant promoting effect on cost stickiness, while debt financing constraints and equity financing constraints have a significant restraining effect on cost stickiness. The internal control quality has a moderation effect on this relationship. In a firm with low quality of internal control, internal financing constraints can enhance cost stickiness, but the weakening effect of external financing on cost stickiness is not affected by internal control quality.

LSTM and GRU Neural Networks as Models of Dynamical Processes Used in Predictive Control: A Comparison of Models Developed for Two Chemical Reactors

This work thoroughly compares the efficiency of Long Short-Term Memory Networks (LSTMs) and Gated Recurrent Unit (GRU) neural networks as models of the dynamical processes used in Model Predictive Control (MPC). Two simulated industrial processes were considered: a polymerisation reactor and a neutralisation (pH) process. First, MPC prediction equations for both types of models were derived. Next, the efficiency of the LSTM and GRU models was compared for a number of model configurations. The influence of the order of dynamics and the number of neurons on the model accuracy was analysed. Finally, the efficiency of the considered models when used in MPC was assessed. The influence of the model structure on different control quality indicators and the calculation time was discussed. It was found that the GRU network, although it had a lower number of parameters than the LSTM one, may be successfully used in MPC without any significant deterioration of control quality.

Do Chief Audit Executives Matter? Evidence from Turnover Events

We investigate the chief audit executive’s (CAE) internal audit supervisory role by examining the change in internal audit monitoring effectiveness following the turnover of CAEs. Using a sample of firms listed on the small and medium enterprise board of China’s stock exchange, we find that CAE turnover is accompanied by a reduction in financial reporting/internal control quality and that the reduction is more pronounced for firms whose successor CAEs have lower financial expertise than their predecessors. Further analysis shows that the negative association with financial reporting/internal control quality is stronger when the turnover is for personal reasons than when it is for internal transfer of the CAE. These findings are robust to a battery of sensitivity checks, including placebo tests and matching diagnostics. Our results highlight the importance of the CAE for a firm’s internal audit functions.