We present a representation learning framework for financial time series forecasting. One challenge of using deep learning models for finance forecasting is the shortage of available training data when using small datasets. Direct trend classification using deep neural networks trained on small datasets is susceptible to the overfitting problem. In this paper, we propose to first learn compact representations from time series data, then use the learned representations to train a simpler model for predicting time series movements. We consider a class-conditioned latent variable model. We train an encoder network to maximize the mutual information between the latent variables and the trend information conditioned on the encoded observed variables. We show that conditional mutual information maximization can be approximated by a contrastive loss. Then, the problem is transformed into a classification task of determining whether two encoded representations are sampled from the same class or not. This is equivalent to performing pairwise comparisons of the training datapoints, and thus, improves the generalization ability of the encoder network. We use deep autoregressive models as our encoder to capture long-term dependencies of the sequence data. Empirical experiments indicate that our proposed method has the potential to advance state-of-the-art performance.

In this paper, we consider the problem of finding a safety criteria between neighboring heavy duty vehicles traveling in a platoon. We present a possible framework for analyzing safety aspects of heavy duty vehicle platooning. A nonlinear underlying dynamical model is utilized, where the states of two neighboring vehicles are conveyed through radar information and wireless communication. Numerical safe sets are derived through the framework, under a worst-case scenario, and the minimum safe spacing is studied for heterogenous platoons. Real life experimental results are presented in an attempt to validate the theoretical results in practice. The findings show that a minimum relative distance of 1.2 m at maximum legal velocity on Swedish highways can be maintained for two identical vehicles without endangering a collision. The main conclusion is that the relative distance utilized in commercial applications today can be reduced significantly with a suitable automatic control system.

This paper develops a controller synthesis method for distributed LQG controlproblems under output-feedback. We consider a system consisting of three interconnected linearsubsystems with a delayed information sharing structure. While the state-feedback case haspreviously been solved, the extension to output-feedback is nontrivial as the classical separationprinciple fails. To nd the optimal solution, the controller is decomposed into two independentcomponents: a centralized LQG-optimal controller under delayed state observations, and a sumof correction terms based on additional local information available to decision makers. Explicitdiscrete-time equations are derived whose solutions are the gains of the optimal controller.

We consider the problem of optimal distributed control with delayed information sharing over chain structures motivated by applications of heavy duty vehicle platooning. We introduce a novel approach to find distributed controllers that take into account communication delays and correlation between the dynamic interconnection. The design decomposes the solution into separate optimization problems. The results show that a tight control is achieved, even in the presence of delays, and behaves well with respect to the imposed disturbances. The computed theoretical cost of the proposed optimal controller is significantly better than the theoretical cost for the centralized control with a global delay and close to the cost for a fully centralized control with full state information at all times.

We consider optimal distributed controller synthesis for an interconnected system subject to communication constraints, in linear quadratic settings. Motivated by the problem of finite heavy duty vehicle platooning, we study systems composed of interconnected subsystems over a chain graph. By decomposing the system into orthogonal modes, the cost function can be separated into individual components. Thereby, derivation of the optimal controllers in state-space follows immediately. The optimal controllers are evaluated under the practical setting of heavy duty vehicle platooning with communication constraints. It is shown that the performance can be significantly improved by adding a few communication links. The results show that the proposed optimal distributed controller outperforms a suboptimal controller in terms of control input energy.

This technical note considers linear quadratic team decision problems. It shows that linear decisions are optimal and can be found by solving a linear matrix inequality.

In this paper, a decentralized control approach for a class of large-scale interconnected systems is proposed. A set of local decentralized controllers is constructed to guarantee the stability of the global system utilizing the notion of superstability. The control scheme and design are based on local state information and model knowledge of the adjacent subsystems. The approach leads to a computationally ecient method where the design procedure is performed on the subsystem level using linear programming. The solutions have been optimized and it is shown that the scheme has some useful features such as monotonic convergence and robustness under interval matrix uncertainty.

In this paper, we investigate the topology convergence problem for the gossip-based Gradient overlay network. In an overlay network where each node has a local utility value, a Gradient overlay network is characterized by the properties that each node has a set of neighbors containing higher utility values, such that paths of increasing utilities emerge in the network topology. The Gradient overlay network is built using gossiping and a preference function that samples from nodes using a uniform random peer sampling service. We analyze it using tools from matrix analysis, and we prove both the necessary and sufficient conditions for convergence to a complete gradient structure, as well as estimating the convergence time. Finally, we show in simulations the potential of the Gradient overlay, by building a more efficient live-streaming peer-to-peer (P2P) system than one built using uniform random peer sampling.

This paper considers linear quadraticteam decision problems where the players inthe team affect each others information structurethrough their decisions. It shows that linear decisionsare optimal and can be found by solving alinear matrix inequality.

It is fuel efficient to minimize the relative distance between vehicles to achievea maximum reduction in air drag. However, the relative distance can only be reduced to acertain extent without endangering a collision. Factors such as the vehicle velocity, the relativevelocity, and the characteristics of the vehicle ahead has a strong impact on what minimumrelative distance can be obtained. In this paper, we utilize optimal control and game theory toestablish safety criteria for heavy duty vehicle platooning applications. The derived results showthat a minimum relative distance of 1.2m can be obtained for two identical vehicles withoutendangering a collision, assuming that there is no delay present in the feedback system. If aworst case delay is present in the system, a minimum relative distance is deduced based uponthe vehicle’s maximum deceleration ability. The relative distance can be reduced if the followervehicle has a greater overall braking capability, which suggests that vehicle heterogeneity andorder has substantial impact. The findings are verified by simulations and the main conclusion isthat the relative distance utilized in commercial applications today can be reduced significantlywith a suitable advanced cruise control system.