Patent Application
20080155293
None
The present invention relates generally to apparatus, systems, and methods for power protection, and more specifically, to apparatus, systems, and methods for validating decision making mechanisms within a power protection system.
Power transmission and distribution networks require an extremely high degree of reliability. Failures in such systems can lead to blackouts. Electrical switchgear, such as circuit breakers and reclosers, are deployed in power networks to isolate faults while maintaining power to as many end users as possible. Usually, a digital intelligent electronic device, such as a protective relay or recloser control, controls the operation of electrical switchgear. However intelligent electronic devices are susceptible to errors caused by background radiation. In particular, memory components used within intelligent electronic devices are susceptible to bit errors caused by high energy particles such as alpha particles or neutrons.
A number of techniques have been disclosed in the prior art detailing methods for reducing errors caused by radiation and other unpredictable sources of errors. For instance, U.S. Pat. No. 6,886,116, issued to Christopher MacLellan, discloses a system for validating error detection logic in a data storage system. MacLellan utilizes a plurality of fault injectors to create erroneous conditions, and then utilizes additional logic to ensure that the error detection logic picks up on the error and does not interfere with the normal operation of the device. MacLellan is a good example of an error detection technique applied to a combined hardware/software system.
U.S. Pat. No. 6,594,774, issued to Craig Chapman and Mark Moeller, focuses exclusively on software errors. In addition to other techniques, Chapman applies the concept of a watchdog timer to individual software processes. A watchdog timer is a hardware timer coupled to a microprocessor that must be reset within a given time period or the watchdog timer causes the microprocessor to reset. In Chapman, individual executable fibers (i.e.; threads or processes) register with a watchdog thread. The executable fibers must then notify the watchdog thread periodically, or the watchdog thread takes a containment action, such as terminating the thread.
Many techniques suited to other industries are not necessarily well suited to the power protection industry. Power protection devices often operate in hostile environments, with large amounts of electromagnetic radiation present. Historically, the power protection industry has dealt with this problem through the use of shielding, grounding, and other basic mechanical and electrical techniques, as well as readback validation of memory structures. Given the sensitivity of the power grid to failures, there is a continuing need within the power protection industry to devise techniques to further reduce the failures of power protection devices and thereby improve the reliability of the power grid.
Accordingly, it is an object of this invention to provide reliable power system automation and control capable of detecting and correcting a large percentage of would-be failures, and thereby raising the overall reliability of the power grid.
Another object of this invention is to provide a system for reliably identifying and isolating faults in a monitored power line with fault detection logic that can, in a large percentage of cases, detect when it has erroneously detected a fault, and prevent the system from taking adverse action based on the erroneously detected fault.
The disclosed invention achieves its objectives through the use of a test data channel that is processed by the same hardware and software that monitors actual data channels. A plurality of input channels containing digital information regarding the power system or a subpart of the power system is analyzed by a computation engine concurrently with a test data channel in a time interleaved manner using the same execution path on the different channels on a sample by sample basis, where an execution path consists of hardware, firmware and software used to implement the utilized algorithm. The test data channel is preferably designed to fully exercise the executable code of the computation engine. The results of the computation performed on the test data channel is compared against a predetermined test result, and if the two differ, a corrective action is taken. Corrective actions range from shutting the system down, to discarding all results from the sampling period when the erroneous reading occurred, to logging the erroneous computation.
The disclosed invention may also be implemented as a method for reliable power system automation or control. A plurality of digital input channels is monitored along with a test data channel. A computation engine performs analysis on all of the channels using a power system analysis algorithm, and produces a plurality of decision data channels and a test result data channel. The test result data channel is compared against a predetermined test result and a corrective action is initiated if the two differ.
Although the characteristic features of this invention will be particularly pointed out in the claims, the invention itself, and the manner in which it can be made and used, can be better understood by referring to the following description taken in connection with the accompanying drawings forming a part hereof, wherein like reference numerals refer to like parts throughout the several views and in which:
Many power control and automation systems process time interleaved channels of data. For instance, a single processor often analyzes three separate phases of current and voltage. An embodiment of the disclosed invention improves the reliability of such a system through the use of additional channels of known test data. The channels of known test data are processed by the same computation engine that processes current and/or voltage data, and a test result channel is produced. The test result channel is then compared against a predetermined test result, and if the test result channel does not match the predetermined test result, an error is noted, and a corrective action is taken.
Referring to the Figures, and in particular to
The computation engine 530 contains executable memory 532, which is susceptible to corruption from external sources such as radiation. Additionally, the computation engine 530, along with any of the clock 514, stimulation source 524, control 525, mux 520, and test logic 544 can be susceptible to corruption from external sources such as radiation. The executable memory 532 is loaded from non-volatile memory 562. Note that many other configurations of memory can utilize the principles of this invention. For instance, the computation engine 530 could execute directly from the non-volatile memory 562, or the non-volatile memory 562 or execution memory 532 and the computation 530 could both be implemented within an FPGA. In this case the execution memory 532 is representative of the configuration memory internal to the FPGA. In some cases the non-volatile memory 562 is not included with the system or represents a mass storage magnetic media.
The computation engine produces trip outputs 536A-C and test result 538 by using the same fault detection algorithm on the different channels of data it receives from multiplexer 520. The test result 538 is then examined by test logic 544, which compares the test result 538 against a predetermined test result. If the test result 538 does not match the predetermined test result a corrective action 560 is output. As depicted, the corrective action 560 is for the non-volatile memory 562 to refresh the executable memory 532 of the computation engine. A number of different techniques may be applied when refreshing the executable memory; among them, the entire executable memory 532 may be refreshed (effectively resetting the computation engine 530), a predetermined section or sections may be refreshed, or sections identified as corrupted may be refreshed. Other potential corrective actions as identified earlier may be taken as well.
Note that the invention described herein utilizes a digital processor. As the algorithms described do not require any particular processing characteristics, any type of processor will suffice. For instance, microprocessors, microcontrollers, digital signal processors, field programmable gate arrays, application specific integrated circuits (ASIC) and other devices capable of digital computations are acceptable where the terms processor or computation engine are used.
The foregoing description of the invention has been presented for purposes of illustration and description, and is not intended to be exhaustive or to limit the invention to the precise form disclosed. The description was selected to best explain the principles of the invention and practical application of these principles to enable others skilled in the art to best utilize the invention in various embodiments and various modifications as are suited to the particular use contemplated. It is intended that the scope of the invention not be limited by the specification, but be defined by the claims set forth below.
