As the power industry moves towards more active distribution networks there is an increased requirement for greater analysis and observability of the current state of the network. There are a number of challenges for utilities in realising this including the quality and accuracy of their network models; the lack of integration between network models and the large quantities of sensor data being collected; the security and communication challenges posed when installing large numbers of sophisticated sensors across distribution networks; and the exponential increase in computing power required to fully analyse modern network configurations. This thesis will look at these challenges and how cloud computing can be used to provide novel solutions by providing secure platforms on which to deploy complex data collection and network analysis applications. One of the main research contributions is the use of remote data collection from Micro Phasor Measurement Units (μPMUs), which collect synchronised information about the state of the distribution network. Impedance equations are applied to network data recorded from μPMUs and the results are compared to network models. This identifies areas of the distribution network as requiring resurveying or upgrading, potentially impacting planning for installation of generation or load. Triggers can be used to reduce the bandwidth of data being sent by a μPMU; these were tested with real world data to highlight how a combination of local intelligence and cloud-based analysis can be used to reduce bandwidth requirements while supporting the use of detailed measurement data for cloud-based analysis in a fault detection system. Power flow analysis is an important tool for both operations and planning engineers, and as computing power has increased the time required to run individual power flow analysis cases has decreased rapidly. However there has also been a corresponding increase in the complexity of the data as utilities seek to model and analyse distributed energy resources attached on the medium and low voltage networks. This has made network models more complex, exponentially increasing the number of contingencies that need to be analysed in an emergency situation. Another main research contribution is a demonstration of the challenges faced when using a commercial cloud platform to inexpensively solve computationally intensive power flow problems and the time, costs and feasibility of performing N-1 and N-2 analysis on a 21,000-bus network. It includes a full analysis and comparison of execution times and costs for different commercial cloud system configurations as well as the extrapolated costs required to run a full N-2 analysis of over 420 million contingencies in under 10 minutes. This includes a demonstration of a cloud client and server application developed as part of this research that leverages a commercial power flow engine. Finally, this thesis will summarise how each of these research outputs can be combined to provide utilities with a commercial, open, standards-based cloud platform for continuous, automated contingency analysis using real-time sensor data based on current network conditions. This would better inform control engineers about areas of vulnerability and help them identify and counter these in real-time.

The master thesis project has been executed with the cooperation of Borlänge Energi, with the aim of reducing the high electric energy dependency which Moldova has on Ukraine, Transnistria and Russia. The project examines what reduction that would be possible by wind power installations on the existing electrical grid of Moldova. The installations should not surpass the capacity of the transmission lines or the voltage levels according to regulation. The southern regions of Moldova proved to have the best wind conditions and the locations of Besarabeasca, Zarnesti, Leovo, Ciadyr and Cimislia in the southern region were chosen for wind power installations. For the analysis a model over the Moldovan electrical system is constructed. Each of the five chosen locations is modelled with a generator symbolizing the wind power installation. The power flow software PSS/E is used to construct the model. To examine possible wind power installations different scenarios are created. The scenarios are executed with the southern regions 110 kV system as a focus area. All scenarios are analysed with a contingency analysis, where transmission lines in the focus region are tripped. The contingency analysis and the scenarios are automated using the programming language Python. An economic analysis shows payback periods for wind power investments in Moldova, the analysis also shows the sensitivity of the electricity price and discount rates. The project concludes that wind power installations are possible with the Moldovan electric grid as it looks today. The installations would result in reducing the high dependency of imported electrical energy.

Submitted by TAISA SCARPIN GUAZI null (taisa_guazi@hotmail.com) on 2017-08-15T20:50:00Z No. of bitstreams: 1 Dissertação Taísa.pdf: 1306309 bytes, checksum: 4293392198ac4c50c1741810879ff2d1 (MD5) / Approved for entry into archive by Monique Sasaki (sayumi_sasaki@hotmail.com) on 2017-08-22T17:58:46Z (GMT) No. of bitstreams: 1 guazi_ts_me_bauru.pdf: 1306309 bytes, checksum: 4293392198ac4c50c1741810879ff2d1 (MD5) / Made available in DSpace on 2017-08-22T17:58:46Z (GMT). No. of bitstreams: 1 guazi_ts_me_bauru.pdf: 1306309 bytes, checksum: 4293392198ac4c50c1741810879ff2d1 (MD5) Previous issue date: 2017-07-28 / Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP) / Em estudo anterior, Guazi e Laurenti identificaram a presença de reforçamento natural na manutenção do comportamento acadêmico de pesquisadores seniores. Contudo, as atuais contingências acadêmicas destoam, em muitos aspectos, das contingências às quais os pesquisadores seniores estavam expostos no início de suas carreiras. As atuais contingências acadêmicas estão permeadas, por exemplo, pelos fenômenos (relativamente recentes) do produtivismo acadêmico e da competitividade entre pesquisadores. Nesse sentido, os professores iniciantes, que ingressaram contemporaneamente na academia, parecem constituir uma geração de pesquisadores cujo perfil é diferente. Considerando esses aspectos, esta pesquisa objetivou identificar as diferenças e similaridades existentes entre as variáveis que controlam o comportamento acadêmico de professores seniores e iniciantes. Participaram desta pesquisa, de natureza exploratória, seis professores bolsistas de produtividade do CNPq que atuam nas áreas de Psicologia, Genética e Física – sendo selecionado, de cada área, um pesquisador sênior e um pesquisador iniciante. Os dados, coletados por meio de entrevistas, foram analisados pelo método de interpretação analítico-comportamental e organizados de modo a permitir a análise, em separado, das contingências originárias e mantenedoras do comportamento acadêmico dos participantes. As três participantes seniores deste estudo são mulheres brancas, que concluíram o doutorado na década de 1960 (os dados coletados permitiram considerar a variável gênero na análise do comportamento acadêmico das participantes seniores), e os participantes iniciantes, são três homens brancos, que se doutoraram entre 2009 e 2012. As contingências responsáveis por instalar o comportamento acadêmico das participantes seniores foram, preponderantemente, contingências de reforçamento positivo, as quais expuseram as entrevistadas à formação liberal. O padrão comportamental acadêmico dos participantes iniciantes, por outro lado, parece ter sido instalado por meio do emprego de regras. Na contemporaneidade, o comportamento acadêmico das participantes seniores parece ser mantido por meio de reforçamento positivo natural. Em contrapartida, as contingências responsáveis por manter o comportamento acadêmico dos participantes iniciantes, as quais estão relacionadas às atuais exigências de produtividade científica, parecem distanciar o comportamento acadêmico dos entrevistados das suas consequências naturais. As principais diferenças encontradas entre as contingências originárias e mantenedoras do comportamento acadêmico dos participantes seniores e iniciantes parecem aludir às diferentes contingências a que os entrevistados foram expostos ao longo da formação básica e superior, e às mudanças pelas quais a educação e ciência brasileira passaram nas últimas décadas. Este estudo contribuiu para a compreensão das práticas científicas contemporâneas, bem como permitiu elucidar as contribuições da análise do comportamento para esta temática. A interpretação dos dados coletados sugere a necessidade de que mais pesquisas, que versam sobre as atuais contingências institucionais acadêmicas, sejam feitas. / In previous study, Guazi and Laurenti identified the presence of natural reinforcement in maintaining the academic behavior of senior researchers. However, current academic contingencies are in many ways at odds with the contingencies to which senior researchers were exposed early in their careers. The current academic contingencies are permeated, for example, by the (relatively recent) phenomena of academic productivism and competitiveness among researchers. In this sense, the beginner professors, who entered the academy contemporaneously, seem to constitute a generation of researchers whose profile is different. Taking these matters into consideration, this research aimed to identify the differences and similarities between the variables that control the academic behavior of senior and beginner professors. Participated in this exploratory research, six academic professors with productiveness scholarships granted by the CNPq who work in the fields of Psychology, Genetics and Physics - a senior researcher and a beginner researcher were selected from each area. The data, collected through interviews were analyzed by the Behavior Analysis interpretation method and organized to allow the analysis, separately, of the contingencies that originated and maintained the participants' academic behavior. The three senior participants in this study were white women, who completed their doctorate in the 1960s (the data collected allowed the gender variable to be considered in the analysis of the academic behavior of senior participants), and the beginner participants were three white men, who completed their doctorate between 2009 and 2012. The contingencies responsible for installing the academic behavior of the senior participants were, predominantly, contingencies of positive reinforcement, which exposed the interviewees to the liberal formation. The academic behavioral pattern of beginner participants, on the other hand, seems to have been installed through the use of rules. Currently, the academic behavior of senior participants seems to be maintained through natural positive reinforcement. On the other hand, the contingencies responsible for maintaining the academic behavior of the beginner participants, which are related to the current demands of scientific productivity, seem to distance the academic behavior of the interviewees from their natural consequences. The main differences found between the contingencies that originated and maintained the academic behavior of the senior and beginner participants seem to allude to the different contingencies to which the interviewees were exposed during basic and higher education, and the changes that Brazilian education and science have undergone in recent years decades. This study contributed to the understanding of the contemporary scientific practices, as well as allowed to elucidate the contributions of the analysis of the behavior to this theme. The interpretation of the data collected suggests the need for more studies on current academic institutional contingencies. / FAPESP: 2015/09081-9

Made available in DSpace on 2014-06-11T19:29:01Z (GMT). No. of bitstreams: 0 Previous issue date: 2011-02-23Bitstream added on 2014-06-13T18:35:24Z : No. of bitstreams: 1 luciano_es_me_bauru.pdf: 769426 bytes, checksum: ab8e48b2e61526d706765a3263da5d66 (MD5) / Vários estudos publicados (IWATTA et al., 2000; MYERS e HOLLAND, 2009; ALMEIDA, 2009; CERQUEIRA, 2009; TAVARES, 2009) demonstram a importância e a possibilidade do ensino de repertórios de análise de contingências para professores, a saber, de repertórios definidos pela execução de interpretações funcionais de interações em sala de aula. Este estudo teve por objetivo o ensino de tais repertórios compreendendo a identificação de respostas que definem queixas escolares, das condições antecedentes e de eventos subsequentes a tais respostas. Participaram deste estudo duas licenciadas em Pedagogia que atuavam como estagiárias no Ensino Fundamental e contratadas por uma Diretoria Municipal de Educação. Foram utilizados protocolos preenchidos por professores da rede municipal. Primeiramente, tais protocolos de alunos dos anos iniciais são encaminhados para o Serviço de Psicologia da referida Diretoria de Educação. O preenchimento foi efetuado antes da realização desta pesquisa, por professoras efetivas da rede municipal dos anos iniciais do Ensino Fundamental. A coleta de dados foi realizada em uma única sessão, na escola em quie as graduandas cumpriam período do estágio. Foram selecionados nove protocolos de encaminhamento para os quais foram construídos modelos de interpretação funcional mediante identificação da resposta (queixa) e dos possíveis eventos antecedentes e subsequentes a tais respostas para cada protocolo. Também foram elaborados dois questionários que indagavam sobre dimensões dos protocolos já preenchidos. o procedimento foi composto por quatro fases. Na fase 1 ocorreu a caracterização inicial de repertórios verbais das participantes na descrição dos Protocolos 1 e 2. Na fase 2, as participantes foram expostas aos modelos de interpretação funcional para os dois protocolos da fase anterior. na Fase 3 foi utilizado um procemento... / Many published essays IWATTA et al., 2000; MYERS e HOLLAND, 2009; ALMEIDA, 2009; CERQUEIRA, 2009; TAVARES, 2009) show the importance and possibilities on teaching contingency analyses repertoires to teachers, which is definite by doing functional interaction interpretations on classroom. This essays aimed to teach those repertoires with proposal identification on school complains about early conditions and subsequent responsible events. Two Education trainnes attended to it, they worked on Elementary school of an educational discrict Directory. It was used formularies, which was fulfilled by the teachers, about their students. First, those formularies were sent to Psychology center and their fulfilling was done by elementary school oficial teachers, before this research was made. This information was taken in one section at school in which trainees work. Nine formularies were selected, which were made functional interpretations, by identifying complain responses and their possible early and late events. Two questionnaires about the questions reach were also made. The procedures were formed by fours sections. The first one, was the initial identifying of participants' vocabulary repertoires to describe formularies 1 and 2. On the second stage, it was given the previous formularies to the participations. on the third stage, it was made a fading out procedure, in order to set grade describing patterns to formularies 3,4, 5 and 6. On fourth stage, a functional emergency words evaluation was made, about possible functional relationship with formularies 7,8 and 9. on the first stage, it was realized that participant analyzed teachers' answers on formularies in ana unexpected way from functional interpretation. The results of fourth stage, in which participants hit 81,26 and 78% in same order, suggest some relevant possible, instructions on procedures adopted on 2 and 3 stages... (Complete abstract click electronic access below)

abstract: Reliable and secure operation of bulk power transmission system components is an important aspect of electric power engineering. Component failures in a transmission network can lead to serious consequences and impact system reliability. The operational health of the transmission assets plays a crucial role in determining the reliability of an electric grid. To achieve this goal, scheduled maintenance of bulk power system components is an important activity to secure the transmission system against unanticipated events. This thesis identifies critical transmission elements in a 500 kV transmission network utilizing a ranking strategy. The impact of the failure of transmission assets operated by a major utility company in the Southwest United States on its power system network is studied. A methodology is used to quantify the impact and subsequently rank transmission assets in decreasing order of their criticality. The analysis is carried out on the power system network using a node breaker model and steady state analysis. The light load case of spring 2019, peak load case of summer 2023 and two intermediate load cases have been considered for the ranking. The contingency simulations and power flow studies have been carried out using a commercial power flow study software package, Positive Sequence Load Flow (PSLF). The results obtained from PSLF are analyzed using Matlab to obtain the desired ranking. The ranked list of transmission assets will enable asset managers to identify the assets that have the most significant impact on the overall power system network performance. Therefore, investment and maintenance decisions can be made effectively. A conclusion along with a recommendation for future work is also provided in the thesis. / Dissertation/Thesis / Masters Thesis Electrical Engineering 2019

Transmission line is a crucial role in power transmission network which connects generating units to consumers. Some unpredicted failure events such as lightening or system faults can cause transmission line tripped, which may bring about a large interruption to the system and causes damage. When line outage happens, the power flow on the tripped line will be redistributed to the rest of lines in the system. It may cause risk of overload happens on other lines, and results in a cascading failure and system collapse. Reasonably, a single line outage will not affect all other lines in the system. Therefore, when a line outage happens, it is important for the system operator to have a preview of which lines will have serious impact and which lines will not, so that the operator can only focus on monitoring certain lines which will be seriously affected, rather than keeping monitoring the whole system. In this thesis, A Line Outage Distribution Factor (LODF) method is proposed and implemented in the IEEE 118 bus system to estimate active power flow redistribution after a line outage. After that, a definition of Thevenin electrical distance between two transmission lines is derived and applied to calculate electrical distances between the outage line and each line in the system. An exponential convergence tendency is found between maximum possible LODF predicted power flow variations and electrical distance, and an exponential regression method is applied to analyze this tendency. The contribution of this work is a rule has been found that starting from the outage line, the maximum possible active power flow variation on transmission lines exponentially decreases exponentially while the electrical distance increases. With only the information of system's normal operating condition and topological information, the maximum possible active power change on each lines caused by single line outage, and the margin of the impact of single line outage on power flow variations may propagate along electrical distance can be easily and quickly predicted. Ultimately, the goal of this work is to allow operators at the control center can concentrate on lines within a certain electrical distance instead of keeping monitoring the whole system when a line outage happens. / Master of Science

Yes / This paper proposes a novel method for planning active distribution networks (ADNs) with the integration of an active network management (ANM) scheme using coordinated voltage control (CVC) through on-load tap changer (OLTC) transformers. The method was formulated as a security-constrained optimal power flow (SCOPF) problem to minimize total operational costs, which maximizes the utilization of renewable distributed generators (DGs) over a planning horizon. The ANM scheme was applied using OLTC to ensure safe operation and reduce voltage violations in the network. To analyse the impact of ANM, the planning problem was examined both with and without the ANM scheme. Moreover, SCOPF, considering the N-1 line contingency analysis and multi-DG configuration, was implemented to analyse the feasibility of the proposed method and the advantages of ANM under contingency situations. The method was validated on a weakly-meshed 16-bus UK generic distribution system (UKGDS). The results showed that ANM can lower operational costs and maintain network voltage for operation in feasible conditions even in the case of a contingency. Moreover, the ANM scheme mitigated the voltage rise effect caused by DGs and maximized their utilization.

O presente trabalho teve por objetivo verificar experimentalmente (1) se a variabilidade do comportamento, obtida como função de contingências de reforçamento LAG n e RDF, sofre interferência da manipulação sistemática dos parâmetros n(LAG) e do limiar de reforçamento (RDF) e (2) se a história de reforçamento, estabelecida por ordem crescente, decrescente e alternada da exigência de variação afeta o grau de variabilidade obtido. Ratos (n=48) machos, da linhagem Wistar, foram divididos em 12 grupos expostos a uma fase de linha de base (FR4) e quatro fases experimentais, nas quais foram manipuladas as duas contingências de reforçamento da variação LAG (com n 2, 4, 8 e 12) ou RDF (com limiar de 1/4, 1/8, 1/12 e 1/16), e duas contingências a essas acopladas, sem exigência de variação. Sob cada contingência, três subgrupos diferiram entre si em relação à ordem de apresentação dos valores de n e limiar, que podia ser crescente (C), decrescente (D) ou aleatória (A). A unidade comportamental foi a sequência de quatro respostas de pressão a duas barras (esquerda e direita). O efeito dessas manipulações foi analisado sobre diferentes variáveis dependentes: índice U, % reforçamento, R/min, tempo de recorrência, frequência por oportunidade, %RE, % repetição e número de alternações.Os dados obtidos mostraram que: (1) os graus de variabilidade são diretamente modificados pelos valores de n ou do limiar; (2) a ordem de apresentação da contingência produziu efeito diferenciado entre os grupos; (3) a simples intermitência de reforços não alterou padrões de variação; (4) as diferentes variáveis dependentes analisadas em conjunto permitiram uma compreensão mais ampla do comportamento de variar / The goal of the present work was to (1) investigate whether operant variability can be influenced by previous contingencies or whether it is solely controlled by the current contingencies; (2) analyze the effects of reinforcement intermittency in producing behavior variability; (3) analyze whether other variability measures are affected when the n value of the LAG contingency and the threshold in a threshold (RDF) procedure are manipulated. Fourty eight Wistar rats were thus exposed to LAG, RDF, and their respective yoked conditions. The n values used for the LAG contingency were 2, 4, 8 and 12, and the threshold values used were 1/4, 1/8, 1/12 e 1/16. The behavior unit analyzed was the sequence of four responses (lever presses) based on the spatial localization of each response (left and right). Subjects were divided into four groups (LAG, RDF, and their respective yoked groups - ALAG and ARDF), and were then exposed to a baseline under a fixed-ratio (FR) 4 schedule and four additional phases with the four possible contingencies. In each group, subjects were exposed to different orders of presentation of the contingencies (values: ascending A, descending D, or random R). The results indicated that: (1) changes in the nor threshold value resulted in changes in the level of variability; (2) the order of presentation of the contingencies had different effects between groups; (3) reinforcement intermittency did not affect variability patterns; (4) response cost controlled the level of behavior variability; (5) the new behavior variability measures were efficient for a broader analysis of variability

Há debates sobre quais são os mecanismos responsáveis por mudanças ocorridas em psicoterapias. Pergunta-se se são as técnicas específicas ou as variáveis da relação terapêutica que propiciam os efeitos da terapia. Também tem sido questionado se mudanças comportamentais produzidas pela terapia são modeladas por contingências da relação terapêutica ou são governadas por novas regras produzidas na terapia. Nas pesquisas sobre emissão de regras (orientação) e autorregras (auto-orientação) anteriores a esta pesquisa não foram encontradas variáveis responsáveis por sua emissão. A determinação da utilização da estratégia de orientação ora parecia ser o cliente, ora o terapeuta, ora nenhum deles, ora o tema abordado, e possivelmente uma complexa combinação destas e de outras variáveis. O estudo detalhado das orientações e auto-orientações de 81 sessões de terapia analítico-comportamental foi realizado nesta pesquisa com o objetivo de verificar quais as variáveis responsáveis pela emissão de orientação e auto-orientação em intervenções clínicas comportamentais. Os resultados indicaram que a maioria dos terapeutas emitiu de 40 a 60 orientações nas nove sessões analisadas. Os clientes dos terapeutas independente da experiência apresentaram poucas auto-orientações. O número de orientações diminuiu para menos da metade quando se contaram apenas orientações com conteúdos e funções diferentes, indicando que os terapeutas tendem a \'repetir\' funcionalmente a orientação. Para as auto-orientações houve diminuição de um quarto ao se contarem aquelas com conteúdo e função diferente. O conjunto de terapeutas emitiu mais orientações para ação específica e genérica; e de forma similar os clientes estes terapeutas emitiram mais auto- orientações para ação específica e auto-orientações para ação genérica. Episódios de orientação/auto-orientação foram identificados nas 81 sessões, ocupando em média um terço das sessões dos terapeutas experientes e um quarto das sessões dos terapeutas pouco experientes. E nesses episódios havia diversos tipos de intervenção do terapeuta além da orientação. Orientações são emitidas de modo geral no seguinte contexto: clientes relatam uma situação vivenciada e algumas intervenções do terapeuta ocorrem. Quando clientes mostram dificuldade em assumir responsabilidade, enfrentar e avaliar seus comportamentos há fornecimento de regras pelo terapeuta. Os clientes na maioria das vezes concordam com as orientações recebidas, em um quarto das ocasiões se opõem a ela e em um sexto recebem novas orientações. Dois terços das auto-orientações foram seguidos de aprovação do terapeuta, mas ocorreram também reprovações. As variáveis intervenientes \"temas\", \"motivação\" e \"escolaridade\" foram consideradas e correlacionadas: existe pouca relação entre tema abordado e presença de orientação/auto-orientação; clientes motivados receberam mais orientações que os desmotivados e resistentes; quanto maior a escolaridade do cliente maior o número de auto-orientações. 10% da amostra foram avaliados por um juiz, com índices de concordância juiz-pesquisador satisfatórios indicando validade externa. Futuras pesquisas devem ser realizadas para correlacionar o uso de regras e os resultados das intervenções clínicas comportamentais. / Mechanisms responsible for changes that occur in psychotherapy are subject to debate. The question is whether specific techniques or variables of the therapeutic relationship promote the effects of the therapy. Another question is whether behavioral changesproduced by therapy are shaped by contingencies of the therapeutic relationship or ifthey are governed by new rules produced in therapy. In researches about emission of rules (orientation) and self emitted rules (self-orientation) prior to this research, novariables responsible for their emission were found. The use of the orientation strategy seemed to be determined either by the client, or by the therapist, sometimes by neither, and at times by the theme, and possibly by a complex combination of these and other variables. A detailed study of orientations and self-orientations in 81 sessions of behavior analytic therapy was carried out in this research to ascertain which variables were responsible for the emission of orientation and self-orientation in behavioralclinical interventions. The findings indicated that each therapist issued an average of 40to 60 orientations during their nine sessions. The clients of both experienced and inexperienced therapists presented few self-orientations. The number of orientationsdecreased to less than half when only rules with different contents and functions were counted, indicating that therapists tend to functionally \'repeat\' the rule. The self-orientations decreased one quarter when those with different content and function were considered. The group of therapists issued more orientations for specific and generic action; similarly, the clients of these therapists issued more self-orientations for specific and generic action. Episodes of orientation/self-orientation were identified in the 81 sessions, occupying an average of one third of the sessions of experienced therapists, and one quarter of the sessions of inexperienced therapists. These episodes involved several types of intervention by the therapist besides orientation. Orientations were generally issued within the following context: clients report a situation they have experienced and the therapist makes some interventions. When clients show difficulty in assuming responsibility, facing and evaluating their behavior, the therapist providesrules. Clients usually agree with the rules they receive, but on one quarter of occasions they oppose them and, and in one sixth of occasions they receive new orientations. Two thirds of self-orientations were approved by the therapist, but there were alsodisapprovals. There was little correlation between the theme and the presence of orientation/self-orientation. Motivated clients received more orientation than unmotivated and resistant clients; the higher the client\'s level of education the greaterthe number of self-orientation. Ten percent of the sessions were evaluated by a judge, with satisfactory indices of judge-researcher agreement indicating external validity. Future researches should be conducted to correlate the use of rules and the results of behavioral clinical interventions.

O objetivo desse trabalho foi (1) comparar a variação comportamental controlada por duas diferentes contingências de reforçamento positivo (LAG e RDF), (2) observar se a ordem de exposição a essas contingências interfere na instalação e na manutenção da variabilidade, e (3) verificar como a extinção interfere no padrão de variabilidade e sua reinstalação Treze ratos foram divididos em dois grupos, que diferiam entre si devido à ordem de exposição a duas contingências (LAG 5 e RDF), em um procedimento ABACA, onde A foi a primeira contingência de reforçamento em vigor, B a segunda e C extinção. A variabilidade foi analisada sobre seqüências de quatro respostas de pressão a duas barras. Os resultados mostraram que ambas as contingências produziram variabilidade comportamental, sendo os maiores níveis de variação obtidos em RDF, independente da ordem de exposição. A extinção produziu níveis de variabilidade diferentes de acordo com a história de reforçamento dos sujeitos, com um pequeno aumento da variabilidade nos sujeitos expostos a LAG 5. O padrão na fase final não foi afetado pela extinção, sendo típico da contingência em vigor. Esses resultados indicam uma fina sensibilidade aos graus de exigência de variação das contingências, sem interferência da ordem de exposição ou da extinção, quando há uma contingência estabelecida, porém na ausência de contingência (extinção) há um maior controle da história previa de reforçamento nos sujeitos. / The purpose of the present work was to (1) compare the behavioral variability controlled by two different contingencies of positive reinforcement (LAG and RDF), (2) examine whether the order of these contingencies interferes on variability acquisition and maintenance, and (3) establish how extinction intervenes in the pattern of variability and its reacquisition. Thirteen rats were divided into two groups that differed from each other in relation to the order of exposure to two contingencies (LAG 5 and RDF) in an ABACA procedure, where A stood for the first contingency of reinforcement in effect, B for the second and C for the extinction. The variability was evaluated over sequences of four press responses on two levers. The results reveal that both contingencies produced behavioral variability, with the highest levels of variation detected in RDF, whichever the order of exposure. The extinction generated different levels of variability according to the subjects\' reinforcement history, with a small increase in variability in subjects exposed to LAG 5. The pattern in the latest phase was not affected by extinction, being typical of the contingency in effect. These results suggest an acute sensitivity to the variability requirement levels of contingencies, without interference of exposure order or extinction when there is a contingency established. However, in the absence of contingency (extinction) there is a highest control by its previous reinforcement history.