Earticle

디지털 마케팅 캠페인을 수행하는 광고주는 매체 기획(media planning) 단계에서 비용대비 최적의 광고 성과를 만들고자 하며, 이를 위해 광고가 집행되는 광고 매체 별로 투입된 예산 대비 광고 성과의 효율을 측정하고 평가하고자 한다. 이러한 분석이 가능하기 위해서는 디지털 마케팅 캠페인이 집행되는 과정 및 그 결과에 해당하는 광고 성과 데이터가 수집 되어야 한다. 구체적으로, 광고 채널 별로 광고가 얼마나 노출되었는지, 광고에 노출된 사용자가 광고주의 사이트에 유입 되었는지, 방문 후 상품을 구매하거나 컨텐츠를 소비하였는지에 해당하는 광고 성과 데이터가 실시간으로 수집 되어야 가능하다. 디지털 기술 덕분에 수많은 사람들의 행동 이력을 실시간으로 추적하고 기록하는데 소요되는 비용이 크게 낮아졌기 때문에 광고 집행 및 성과에 대한 데이터 수집이 가능하게 되었다 (Goldfarb and Tucker 2019). 실제로 광고 매체 들은 자사의 광고 플랫폼을 통해 광고 데이터를 수집하고 있으며, 광고주를 대상으로 광고 집행 현황과 광고 성과를 요약해서 제공하고 있다. 광고주가 온라인 상에서 디지털 마케팅을 수행할 때 단일 매체에만 의존하기 보다는 시장에 존재하는 다수의 매체 상에 광고 예산을 배분하여 광고를 집행하고자 하기 때문에, 매체 별로 얼마의 광고비를 지출하는 것이 최적의 의사결정인지에 대한 답을 찾기 원한다. 최근 연구에서는 매체 간 사용자들이 중복되어 있다는 점에 착안하여, 예를 들어 A라는 매체와 B라는 매체에서 광고를 보고 상품을 구매한 경우, 두 매체 별로 동일 사용자에게 영향을 미친 기여도를 측정하는 연구가 진행된 바 있으나 (Ghose and Todri 2016), 이러한 연구는 광고 예산 최적화 관점 보다는 기여도 모델을 제시하고 성과를 측정하는 방법을 소개하고 있기 때문에 광고주 입장에서 찾고자 하는 예산 최적화에 대한 구체적인 답을 제공하지 못한다는 한계가 있다. 광고 예산 최적화를 수행해온 대표적인 연구로는 자료포괄분석(Data Envelopment Analysis) 모형이 있다. 자료포괄분석은 비모수적 선형계획법으로, 투입량에 따른 산출 효율성을 다른 의사결정단위와 상대적 관점에서 비교하여 투입량을 줄이거나 늘여야 하는지를 계산하는 방법론이다. 자료포괄분석 모형은 회귀분석과 같은 모수에 기반한 모형과 달리 구체적인 수학적 함수를 가정하지 않기 때문에, 투입량과 산출량 사이의 명확한 인과관계를 밝히기 어려운 의사결정단위들의 상대적인 효율성을 판단할 수 있다는 장점이 있다 (Charnes 외 1979). 그러나 자료포괄분석 모형은 비모수적 선형계획법으로 모수에 기반한 통계적 유의성 검정 과정이 없기 때문에 모형의 판별력을 측정할 수 없는 단점을 가지고 있으며, 최적화 해야하는 목적 함수가 비선형일 경우 적용할 수 없다는 한계를 가지고 있다. 그러나 광고 성과는 한계효용이 체감하는 비선형 함수의 형태를 갖는다는 것이 기본 가정이기 때문에, 목적함수가 선형이라는 가정을 적용할 수 없는 기존 방법론의 한계를 극복하고자 본 연구에서는 비선형 광고성과 목적함수에 대하여 다매체 포트폴리오 상에서 최적의 광고 예산 해를 구하는 모형을 제안한다. 본 연구에서 제안하는 다매체 광고 예산 배정 최적화 모형은 광고주가 최적의 광고 성과를 만들기 위하여 각 매체 별로 얼마의 예산을 배정해야 하는지를 찾아주기 위하여 설계되었다. 매체 별로 집행되는 광고비가 증가할수록 광고 성과는 점차 체감한다는 수확체감의 법칙(law of diminish returns)을 바탕으로 (Broussard 2000; Smith 외 2006; Simon & Arndt 1980), 최적화 문제의 목적함수가 볼록함수가 되도록 로그함수를 가정하여 최적화 문제에 대한 해를 항상 찾을 수 있다는 것에 대한 이론적인 지지를 받고자 하였다. <표 1> 에서 설명하는 목적함수는 광고 예산 최적해를 구하고자 하는 시점(t) 에 대해서, 로그함수를 통해 추정되는 광고 성과의 총합을 의미한다. 이 때, ai, bi 는 i번째 광고매체에 대해 관찰된 모든 과거 시점의 데이터를 사용하여 로그함수를 추정해준 파라미터 값에 해당한다. 또한, 매체별광고예산i,t 은 i번째 광고매체에 t시점에 배정된 예산이며, 동일 시점의 Σi매체별광고예산i,t 은 총광고예산t 보다 작게 설명하였다. 이는 기업의 사업계획서 상에 따른 집행 가능한 광고 예산이 사전에 정의되어 있다는 가정하에 해당 예산 한도 내에서 각 매체 별 광고 예산을 배정하는 점을 반영하였다. 제안한 다매체 광고 예산 배정 비선형 최적화 모형을 기반으로 다음과 같이 네 가지 단계로 구성되는 미디어 플래닝 자동화 프로세스를 설계하였다. 단계1 에서는 광고 매체 별 성과 데이터 및 광고 집행 예산 이력을 집계하여 관찰하고자 하는 기간 동안 광고 예산대비 성과가 얼마나 발생하였는지를 체계적으로 정리한다. 단계2 에서는 광고 예산과 광고 성과 간의 관계를 로그함수로 표현하기 위한 파라미터 추정을 수행한다. 단계3 에서는 <표 1> 에서 제시한 모형의 해를 찾기 위하여 목적함수가 로그함수인 경우에 적용할 수 있는 솔버(solver)를 사용하여 (예: Stanford 대학 Stephen Boyd 교수가 개발한 CVX) 해를 구한다. 단계4 에서는 최적화 결과를 바탕으로 광고 예산을 을 줄여야 하는 매체와, 증가해야 하는 매체를 파악하여 구체적으로 얼마나 광고 예산을 증가시키거나 감소시켜야 하는지를 확인하여 실제 광고주나 광고 에이전시 실무자에게 피드백을 받아 진행하게 된다. 이러한 프로세스를 통해 광고주가 디지털 마케팅 캠페인을 집행하고자 하는 광고 매체 별로 예산을 배정하여 최적의 광고 성과를 내도록 하는 미디어 플래닝을 자동으로 수행하게 되므로, 데이터 분석 기반의 과학적인 광고 의사결정이 가능하도록 기여한다.

This study investigates the optimization of sectional shape with two dimensions on the rubber gasket of electric vehicle battery in order to maintain the airtightness and watertightness. For the section optimization, the shape of protruding section was analyzed as design variables and the design point was composed by the design of experiment(DOE) for the selected protruding shape. The uniaxial tensile test was carried out for the analysis of rubber gasket and five parameters of Mooney-Rivlin hyperelastic model were derived from the test data in order to construct the strain energy function for nonlinear behavior. The rubber gasket compression analysis was performed by using ANSYS of a commercial software and the performance of optimal shape was verified by performing the tests of watertightness and airtightness on the 3D rubber gasket with the derived section.

In the real life learning process, the teacher communicates with the students for a better learning outcome. The teaching-learning-based optimization (TLBO) algorithm simulates this procedure and shows its great performance in solving the constrained and unconstrained nonlinear optimization problem. This paper presents an adaptive direction strategy（ADS ）t o improve the searching ability for the TLBO algorithm. The improved algorithm is tested through searching the optimal points for a few typical testing functions. The testing result shows that the improved TLBO algorithm could obtain better optimal solutions in shorter time. Compared to the normal TLBO algorithm, the stability and effectiveness of the improved algorithm are increased greatly.

Maximize coverage and prolong the network lifetime for wireless sensor networks has become one of the key topics of research. For this, put forward a kind of node scheduling strategy based on covering algorithm. The algorithm by Poisson distribution model structure node density formula, according to the node density formula in the monitoring region deploying to effective coverage; On the other hand, through the node state scheduling mechanism updates as well as to the neighbor node, can be made of residual energy of nodes and monitoring the area sensor the node energy consumption balance, so as to prolong the network lifetime goal. The simulation results show that, the algorithm cans not only useless nodes completely too effective coverage, improve the coverage while optimizing the cyber source configuration, prolong the network lifetime.

Genetic algorithms for multi-objective optimization problem to be solved were studied. Through the elitist strategy analysis, it is an improved multi-objective optimization algorithm. The algorithm uses a data warehouse to store the optimal solution produced by individuals in each generation, from the way individuals adopt measures to phase out the individual data warehouse identical or similar, the algorithm also improved selection operator, so that the algorithm adaptive capacity enhancement, the new algorithm improves the algorithm performance, improves the quality of understanding between sets, can get a lot of optimal and balanced.

The paper describes an optimal parameter selection method of automatic voltage regulator (AVR) for a large power plant in order to improve power system stability. The AVR is generally used to control the output voltage of the generator by regulating the excited voltage in a generator rotor. In order to improve transient stability in power system, the paper proposes the method to select optimal parameters of the AVR based on exact power system modeling. The performances of the proposed method are carried out by time-domain simulation by applying fault to a power system

In the process of coverage for multiple targets, due to the existence of a large number of redundant data make the effective monitoring area coverage decreased and force the network to consume more energy. Therefore, this paper proposes a multi-target k-coverage preservation protocol. First of all, establish the affiliation between the sensor nodes and target nodes through the network model, present a method to compute the coverage expected value of the monitoring area; secondly, in the network energy conversion, using scheduling mechanism in sensor nodes to attain the network energy balance, and achieve different network coverage quality through different nodes energy conversion. Finally, simulation results show that NMCP can effectively reduce the number of active nodes meeting certain coverage requirements and then improve the network lifetime.

In this paper, the search is conducted along each of the coordinate directions for finding the minimum. If $e_i$ is the unit vector along the coordinate direction i, we determine the value a, minimizing f(a)= $f(x+ae_i)$, where a is a real number. A move is made to the new point $x+a_ie_i$ at the end of the search along the direction i. In an n dimensional problem, we define the search along all the directions as one stage. The function value at the end of the stage is compared to the value at the beginning of the stage in establishing the convergence. The gradient appears to be zero at point. We can safeguard this by introducing an acceleration step of one additional step along the pattern direction developed by moves along the coordinate directions.

In this paper, we have proposed a method to solve fuzzy nonlinear optimization problems with inequality constraints problems. It is based on the pseudo-invex concept and Karush-Kuhn-Tucker optimality conditions. By using the $\alpha$-couple level, we have obtained the objective functions with the interval values. In order to focus on this, we have suggested two kinds of optimal solutions according to the partial order defined on closed intervals. With this new approach, we have dealt with five didactic examples. This helped us show how our method works and compare it to other methods. Through a ranking function, we have shown that our method is the best option to solve nonlinear optimization problems with inequality constraints.

In this paper, we construct a new approach of affine scaling interior algorithm using the affine scaling conjugate gradient and Lanczos methods for bound constrained nonlinear optimization. We get the iterative direction by solving quadratic model via affine scaling conjugate gradient and Lanczos methods. By using the line search backtracking technique, we will find an acceptable trial step length along this direction which makes the iterate point strictly feasible and the objective function nonmonotonically decreasing. Global convergence and local superlinear convergence rate of the proposed algorithm are established under some reasonable conditions. Finally, we present some numerical results to illustrate the effectiveness of the proposed algorithm.

The fundamental research for the mobile robot navigation using the numerical optimization method is presented. We propose an image-based visual servo navigation algorithm for a wheeled mobile robot utilizing a ceiling mounted camera. For the image-based visual servoing, we define the composite image Jacobian which represents the relationship between the speed of wheels of a mobile robot and the robot's overall speed in the image plane. The rotational speed of wheels of a mobile robot can be directly related to the overall speed of a mobile robot in the image plane using the composite image Jacobian. We define the mobile robot navigation problem as an unconstrained optimization problem to minimize the cost function with the image error between the goal position and the position of a mobile robot. In order to avoid the obstacle, the modified cost function is proposed which is composed of the image error between the position of a mobile robot and the goal position and the distance between the position of a mobile robot and the position of the obstacle. The performance was evaluated using the simulation.

This paper presents a method of suboptimal control for nonlinear systems via block pulse transformation. The adaptive optimal control scheme proposed by J.P. Matuszewski is introduced to minimize the performance index. Nonlinear systems are controlled using the obtained optimal control via block pulse transformation. The proposed method is simple and computationally advantageous. Viablity of the this method is established with simulation results for the van der Pol equation for comparision with other methods.

Flight trajectory optimization has become an important factor not only to reduce the operational costs (e.g.,, fuel and time related costs) of the airliners but also to reduce the environmental impact (e.g.,, emissions, contrails and noise etc.) caused by the airliners. So far, these factors have been dealt with in the context of 2D and 3D trajectory optimization, which are no longer efficient. Presently, the 4D trajectory optimization is required in order to cope with the current air traffic management (ATM). This study deals with a cubic spline approximation method for solving 4D trajectory optimization problem (TOP). The state vector, its time derivative and control vector are parameterized using cubic spline interpolation (CSI). Consequently, the objective function and constraints are expressed as functions of the value of state and control at the temporal nodes, this representation transforms the TOP into nonlinear programming problem (NLP). The proposed method is successfully applied to the generation of a minimum length optimal trajectories along 4D waypoints, where the method generated smooth 4D optimal trajectories with very accurate results.

In this paper, we focus on maximizing weighted sum energy efficiency (EE) for a multi-cell multi-user channel. In order to solve this non-convex problem, we first decompose the original problem into a sequence of parallel subproblems which can optimized separately. For each subproblem, a base station employs dirty paper coding to maximize the EE for users within a cell while regulating interference induced to other cells. Since each subproblem can be transformed to a convex multiple-access channel problem, the proposed method provides a closed-form solution for power allocation. Then, based on the derived optimal covariance matrix for each subproblem, a local optimal solution is obtained to maximize the sum EE. Finally, simulation results show that our algorithm based on non-linear precoding achieves about 20 percent performance gains over the conventional linear precoding method.

The present study aims to investigate the shimmy stability behavior of a single wheeled nose landing gear system. The system is supposed to be equipped with an electromechanical actuator capable to control the shimmy vibrations. A Proportional-Integrative-Derivative (PID) controller, tuned by using the Particle Swarm Optimization (PSO) procedure, is here proposed to actively damp the shimmy vibration. Time-history results for some test cases are reported and commented. Stochastic analysis is last presented to assess the robustness of the control system.

According to the controllability of pulse times and the amount of jumps in the states at these times in the process of fed-batch culture producing 1,3-propanediol, this paper proposes a terminal optimal control model, whose constraint condition is the nonlinear impulsive delay system. The existence of optimal control is discussed and an optimization algorithm which is applied to each subinternal over one cycle for this optimal control problem is constructed. Finally, the numerical simulations show that the terminal intensity of producing 1,3-propanediol has been increased obviously.

초탄성을 고려한 비선형 구조의 레벨셋 기반 위상 및 형상 최적설계 방법을 개발하였다. 전체 영역에서 재료의 극단적인 불균형 분포로 기인하는 부정확한 접강성행렬(tangent stiffness matrix)로 인해, 비선형 문제의 위상 최적설계는 심각한 수렴성의 어려움을 겪는다. 이를 해결하기 위해, 임의의 형상을 표현할 수 있는 레벨셋 방법의 장점을 이용하여 정확한 접강성 행렬을 구하기 위해 명시적인 경계(explicit boundary)를 이용하였다. 레벨셋 함수로 표현되는 임의의 영역을 암시적 고정 격자(implicit fixed grid)를 이용하여 계산하는 것 대신에 명시적으로 그 영역을 이산화하기 위해 딜라우네이 삼각화 기법(Delaunay triangulation scheme)을 이용하였다. 레벨셋 방정식을 풀기 위해 최적화 조건으로부터 라그란지안(Lagrangian; 목적함수)가 감소하는 방향이 되도록 속도장을 결정하였다. 실제 영역 바깥쪽 속도장은 Adalsteinsson와 Sethian(1999)가 제안한 속도확장 기법을 이용하여 구하였다. 레벨셋 기반의 최적화 기법에 위상 민감도를 이용하여, 최적화 과정에서 원하는 시기와 위치에 위상 변화가 가능하도록 하였다.

A level set based topological shape optimization method for nonlinear structure considering hyper-elastic problems is developed. To relieve significant convergence difficulty in topology optimization of nonlinear structure due to inaccurate tangent stiffness which comes from material penalization of whole domain, explicit boundary for exact tangent stiffness is used by taking advantage of level set function for arbitrary boundary shape. For given arbitrary boundary which is represented by level set function, a Delaunay triangulation scheme is used for current structure discretization instead of using implicit fixed grid. The required velocity field in the actual domain to update the level set equation is determined from the descent direction of Lagrangian derived from optimality conditions. The velocity field outside the actual domain is determined through a velocity extension scheme based on the method suggested by Adalsteinsson and Sethian(1999). The topological derivatives are incorporated into the level set based framework to enable to create holes whenever and wherever necessary during the optimization.

This paper presents a fuzzy set method to enforce complementarity constraints (CCs) in a nonlinear interior point method (NIPM)-based optimization. NIPM is a Newton-type approach to nonlinear programming problems, but it adopts log-barrier functions to deal with the obstacle of managing inequality constraints. The fuzzy-enforcement method has been implemented for CCs, which can be incorporated in optimization problems for real-world applications. In this paper, numerical simulations that apply this method to power system optimal power flow problems are included.

We propose a navigation algorithm using image-based visual servoing utilizing a fixed camera. We define the mobile robot navigation problem as an unconstrained optimization problem to minimize the image error between the goal position and the position of a mobile robot. The residual function which is the image error between the position of a mobile robot and the goal position is generally large for this navigation problem. So, this navigation problem can be considered as the nonlinear least squares problem for the large residual case. For large residual, we propose a method to find the second-order term using the secant approximation method. The performance was evaluated using the simulation.

This paper presents a new approach for the design of robust $H_{\infty}$ sliding mode observer (SMO) for a class of Lipschitz nonlinear systems where both faults and uncertainties are considered. A sufficient condition using linear matrix inequality (LMI) optimization is derived to guarantee the asymptotically stability of the estimation error dynamics and compute the observer gains. A fault estimation scheme is presented where the estimation signal can approximate the fault to some degree of accuracy. Our design approach has some advantages. The Lipschitz constant of the nonlinear term in the system and the disturbance attenuation level are maximized simultaneously through convex multiobjective optimization. For this reason, the Lipschitz constant is suitable to a large class of uncertain nonlinear systems. Moreover, the fault estimation is much more robust against disturbances and nonlinear uncertainty and can preserve the fault signal shape effectively. Finally, a simulation study on a robotic arm system is presented to show the effectiveness of this approach.