Abstract: According to one aspect of the present disclosure, a battery circuit is disclosed that includes a plurality of battery cells. The plurality of battery cells includes at least one connected battery cell and at least one disconnected battery cell. A first switch is configured to disconnect a particular one of the at least one connected battery cells, and a second switch is configured to connect a particular one of the at least one disconnected battery cells. A controller is configured to control the first and second switches based on a detected condition in the particular connected battery cell.

Abstract: According to one aspect of the present disclosure, a battery circuit is disclosed that includes a plurality of battery cells. The plurality of battery cells includes at least one connected battery cell and at least one disconnected battery cell. A first switch is configured to disconnect a particular one of the at least one connected battery cells, and a second switch is configured to connect a particular one of the at least one disconnected battery cells. A controller is configured to control the first and second switches based on a detected condition in the particular connected battery cell.

Abstract: The waveforms (FIGS. 57-80) of signals derived from electrostatic probes (1, 2, 3) are classified (FIG. 9) in terms of width (FIG. 14), presence or absence of following peaks (FIG. 15), presence of multiple peaks (FIG. 21), frequency content (FIG. 22), and noise content of a portion thereof (FIG. 23) as well as polarity utilizing routines related to the various probes (FIGS. 10-12).

Abstract: The wave shapes (FIGS. 57-80) of signals derived from electrostatic probes (1, 2, 3) disposed in the gas stream of a gas turbine engine are classified (FIG. 9) into categories relating to engine and other events, and are further correlated or discriminated by other conditions of the engine. Engine acceleration and the sign thereof (FIG. 4) are utilized to correlate and discriminate acceleration and deceleration (FIG. 81). Power lever angle (68, FIG. 1) is utilized (FIG. 4A) to provide an indication of afterburner being lighted, a flameholder probe (27, FIG. 1) becomes biased (82, 83, 85, FIG. 1) to sense when the afterburner is lit, for correlation with afterburner nozzle liner erosion (FIG. 69, FIG. 81). Chopping the fuel to the afterburner is correlated (FIG. 4A, FIG. 81) with signals indicative of the character of the afterburner chop (FIGS. 77-79).

Abstract: Signals provided by probes (1, 2, 3) disposed in the gas stream of a gas turbine engine are classified in accordance with discriminators (Tables 2-9) utilized in various routines (FIGS. 25-56), including some discriminators which are not currently used, but may allow further discrimination of additional signals, when recognized. The categories of waveshapes (FIGS. 57-80) which are classified includes categories (FIGS. 70 and 71) which are readily distinguishable but have not as yet been correlated to engine events, thereby providing a data history base in the event they become correlated and aiding in the discrimination of these signals from other, correlated signals.

Abstract: The wave shapes of signals derived from electrostatic probes (1, 2, 3) disposed in the gas stream of a gas turbine engine are utilized to correlate the signals (FIGS. 57-59; FIGS. 62-80) with causal engine events, and also to correlate some of such signals (FIGS. 60 and 61) to faults in the diagnostic system, utilizing a special routine (FIG. 44) and also utilizing a routine that can discriminate between fault and engine event (FIG. 45).

Abstract: Records of signals (FIG. 6) derived from probes (1, 2, 3) in a gas turbine engine are captured (FIG. 8) so as to provide digital signal records indicative of waveshapes (FIGS. 57-80) which can be correlated to identify the causal engine event, including history samples which occur before a record-starting threshold crossing (FIGS. 6 and 8).

Abstract: The quality factors (Q) indicative of the likelihood of the waveshape (FIGS. 57-80) of signals produced by electrostatic probes (1, 2, 3, FIG. 1) disposed in the gas stream of a gas turbine engine are combined for certain categories of sets of the probes which will simultaneously sense the same engine event (FIGS. 9, 12a, 12b). The likelihood of signals correlating to particular engine event as a function of positive and negative threshold crossings (FIG. 83) is correlated as between pairs of probes, both digitally (FIGS. 85 and 86) and by analog hardware (FIGS. 87 and 88).

Abstract: The waveforms (FIGS. 57-80) of signals produced by electrostatic probes (1, 2, 3, FIG. 1) are established in records (FIGS. 5 and 6) about 80 milliseconds long; if successive records are sensed contiguously (into N, without any break in-between) the first record of a string is provided with a flag indicative thereof (FIG. 8a) and the second and last records in the contiguous string are stored in special buffers (FIGS. 8b and 8c). All of the other records in-between are thrown away. The string of contiguous records are treated specially: the three records have their characteristics combined in identifying acceleration/deceleration (FIG. 80) and peeled off abraidable seals (FIGS. 62 , FIG. 45c); only the first record is utilized to identify the start of an afterburner chop (FIGS. 77 and 54a) and only the last record is utilized to identify the end of a chop (FIGS. 78, 79 and 54a).

Abstract: The categories assigned to waveshapes (FIGS. 57-80) of signals derived from electrostatic probes (1, 2, 3, FIG. 1) disposed in the gas stream of a gas turbine engine are combined over a classification record group of about 20 or 30 records (FIG. 81). The possible causation of signals identified by positive and negative threshold crossings is statistically combined by summation across a group of records (FIGS. 85 and 86). The quality values indicative of the likelihood of a waveform correlating to a catagory of engine event are combined across records, category by category, before decisions are made (FIGS. 87 and 88).

Abstract: Engine distress is indicated by ten times a normal count for a healthy engine, of signals derived from electrostatic probes (1, 2, 3) disposed in the gas stream of a gas turbine engine, within a variable time interval which is established (FIG. 7) as the integral of R.M.S. noise on one of the probes (3), thereby providing a variable time intervale indicative of the rate at which the engine is working.

Abstract: Equipment which analyzes the wave shapes (FIGS. 57-80) of signals derived from electrostatic probes (1, 2, 3, 27) disposed in the gas stream of a gas turbine engine operates differently in dependence upon different engine conditions. When surge is detected (1, 90-93, FIG. 1) the signal records are stored away specially (FIGS. 8 and 8b) and classified specially (FIGS. 8c, 9, 9a and 9b) when a request therefor (93-95, FIG. 1) is made. Acceleration signals (FIG. 80) are looked for (FIG. 12) only when there is significant engine acceleration (FIG. 4). Only nozzle liner erosion (FIG. 69) is looked for (FIG. 12) when the afterburner is lit (89, FIG. 1), in one probe (3, FIG. 1) while at the same time afterburner rumble (FIG. 80) is looked for in a flameholder probe (27, FIG. 1) when the afterburner is lit. When the afterburner fuel is being chopped (FIG. 4a) only afterburner chop signals (FIGS. 77-79) are looked for (FIG. 12). When the engine speed is below idle, no records of signals are made (FIG. 8).

Abstract: A voltage threshold utilized (FIG. 8) to recognize signals of interest provided by probes (1, 2, 3) of a gas turbine engine for analysis of waveshapes (FIGS. 57-80) indicative of the engine event causing the signals, includes an acceleration component (FIGS. 3 and 4), whereby to avoid collecting records of signals related only to normal acceleration and deceleration, without masking abnormal signals which occur with acceleration or deceleration.