Abstract:
Systems and methods are provided for the control of an industrial asset, such as a power generating asset. Accordingly, a cyber-attack model predicts a plurality of operational impacts on the industrial asset resulting from a plurality of potential cyber-attacks. The cyber-attack model also predicts a corresponding plurality of potential mitigation responses. In operation, a cyber-attack impacting at least one component of the industrial asset is detected via the cyber-attack neutralization module and a protected operational impact of the cyber-attack is identified based on the cyber-attack model. The cyber-attack neutralization module selects at least one mitigation response of the plurality of mitigation responses based on the predicted operational impact and an operating state of the industrial asset is altered based on the selected mitigation response.
Abstract:
An industrial asset may have monitoring nodes (e.g., sensor or actuator nodes) that generate current monitoring node values. An abnormality detection and localization computer may receive the series of current monitoring node values and output an indication of at least one abnormal monitoring node that is currently being attacked or experiencing a fault. An actor-critic platform may tune a dynamic, resilient state estimator for a sensor node and output tuning parameters for a controller that improve operation of the industrial asset during the current attack or fault. The actor-critic platform may include, for example, a dynamic, resilient state estimator, an actor model, and a critic model. According to some embodiments, a value function of the critic model is updated for each action of the actor model and each action of the actor model is evaluated by the critic model to update a policy of the actor-critic platform.
Abstract:
An industrial asset may have monitoring nodes that generate current monitoring node values. A dynamic, resilient estimator may split a temporal monitoring node space into normal and one or more abnormal subspaces associated with different kinds of attack vectors. According to some embodiments, a neutralization model is constructed and trained for each attack vector using supervised learning and the associated abnormal subspace. In other embodiments, a single model is created using out-of-range values for abnormal monitoring nodes. Responsive to an indication of a particular abnormal monitoring node or nodes, the system may automatically invoke the appropriate neutralization model to determine estimated values of the particular abnormal monitoring node or nodes (e.g., by selecting the correct model or using out-of-range values). The series of current monitoring node values from the abnormal monitoring node or nodes may then be replaced with the estimated values.
Abstract:
A flutter control system for a turbine includes a processor. The processor is configured to detect blade flutter of a turbine. The blade flutter indicates that blades of the turbine are in a deflected position different from a nominal operating position. The processor is configured to control operational parameters of the turbine that reduce or eliminate the blade flutter to improve the reliability and efficiency of the turbine.
Abstract:
A method includes selecting a first desired parameter of a machinery configured to produce power, a first surrogate parameter related to the desired parameter, and a first model configured to generate the desired parameter based on a first relationship between the first surrogate parameter and the first desired parameter. The method also includes receiving data related to the first surrogate parameter from a plurality of sensors coupled to the machinery and generating the first desired parameter using the data and the first model. Further, the method includes deriving a first set of empirical data relating the first surrogate parameter to the desired parameter and adjusting the first model based on the data, the first surrogate parameter, and the first set of empirical data, wherein the adjustment to the first model occurs in real-time.
Abstract:
A control system for an adaptive-power thermal management system of an aircraft having at least one adaptive cycle gas turbine engine includes a real time optimization solver that utilizes a plurality of models of systems to be controlled, the plurality of models each being defined by algorithms configured to predict changes to each system caused by current changes in input to each system. The real time optimization solver is configured to solve an open-loop optimal control problem on-line at each of a plurality of sampling times, to provide a series of optimal control actions as a solution to the open-loop optimal control problem. The real time optimization solver implements a first control action in a sequence of control actions and at a next sampling time the open-loop optimal control problem is re-posed and re-solved.
Abstract:
Methods, apparatus, systems, and articles of manufacture are disclosed to perform prognostic health monitoring of a turbine engine. An example apparatus includes a health quantifier calculator to execute a computer-generated model to generate first sensor data of a turbine engine, the first sensor data based on simulating a sensor monitoring the turbine engine using asset monitoring information, a parameter tracker to execute a tracking filter using the first sensor data and second sensor data to generate third sensor data corresponding to the turbine engine, the second sensor data based on obtaining sensor data from a sensor monitoring the turbine engine, the third sensor data based on comparing the first sensor data to the second sensor data, the health quantifier calculator to execute the computer-generated model using the third sensor data to generate an asset health quantifier of the turbine engine; and a report generator to generate a report including the asset health quantifier and a workscope recommendation based on the asset health quantifier when the asset health quantifier satisfies a threshold.
Abstract:
An industrial asset may have monitoring nodes that generate current monitoring node values. An abnormality detection computer may determine that an abnormal monitoring node is currently being attacked or experiencing fault. A dynamic, resilient estimator constructs, using normal monitoring node values, a latent feature space (of lower dimensionality as compared to a temporal space) associated with latent features. The system also constructs, using normal monitoring node values, functions to project values into the latent feature space. Responsive to an indication that a node is currently being attacked or experiencing fault, the system may compute optimal values of the latent features to minimize a reconstruction error of the nodes not currently being attacked or experiencing a fault. The optimal values may then be projected back into the temporal space to provide estimated values and the current monitoring node values from the abnormal monitoring node are replaced with the estimated values.
Abstract:
A control system for an adaptive-power thermal management system of an aircraft having at least one adaptive cycle gas turbine engine includes a real time optimization solver that utilizes a plurality of models of systems to be controlled, the plurality of models each being defined by algorithms configured to predict changes to each system caused by current changes in input to each system. The real time optimization solver is configured to solve an open-loop optimal control problem on-line at each of a plurality of sampling times, to provide a series of optimal control actions as a solution to the open-loop optimal control problem. The real time optimization solver implements a first control action in a sequence of control actions and at a next sampling time the open-loop optimal control problem is re-posed and re-solved.
Abstract:
A system includes an emissions control system. The emissions control system includes a processor programmed to receive one or more selective catalytic reduction (SCR) operating conditions for an SCR system. The SCR system is included in an aftertreatment system for an exhaust stream. The processor is also programmed to receive one or more gas turbine operating conditions for a gas turbine engine. The gas turbine engine is configured to direct the exhaust stream into the aftertreatment system. The processor is further programmed to derive a NH3 flow to the SCR system based on an SCR model and the one or more SCR operating conditions, to derive a NO/NOx ratio, and to derive a fuel split for the gas turbine engine based on the NH3 flow, the NO/NOx ratio, or a combination thereof.
Abstract translation:系统包括排放控制系统。 排放控制系统包括被编程为接收用于SCR系统的一个或多个选择性催化还原(SCR)操作条件的处理器。 SCR系统包括在废气流的后处理系统中。 处理器还被编程为接收用于燃气涡轮发动机的一个或多个燃气轮机操作条件。 燃气涡轮发动机构造成将排气流引导到后处理系统中。 处理器进一步被编程为基于SCR模型和一个或多个SCR操作条件导出到SCR系统的NH 3流,以导出NO / NO x比率,并且基于燃料涡轮发动机导出燃气涡轮发动机的燃料分流 NH 3流量,NO / NO x比率或其组合。