Monitoring system of VRLA battery capacitance
First Claim
1. A method of monitoring the capacity of a valve regulated lead acid battery, comprising:
- connecting at least one battery monitor to said valve regulated lead acid battery;
connecting said monitor to a centralized system through an industry standard data system that collects and transfers data from the field to a central office;
providing said centralized system with an alarm;
performing a short-term discharge test on said battery using said battery monitor; and
obtaining a predicted capacity using a neural network and fuzzy logic network in combination with a prediction algorithm;
training said neural network for a specific battery type in a lab by determining the actual capacity of a battery, inputting parameters with the actual capacity used as a target, and programming the set of coefficients yielded by said neural network in said battery monitor;
wherein, said discharge test provides input parameters for said networks which in combination with said prediction algorithm calculate said predicted capacity, and wherein, said alarm is activated when said predicted capacity falls below a predetermined percentage, when an individual cell voltage is reduced to about 1.95 volts or less, or when a system failure occurs.
0 Assignments
0 Petitions
Accused Products
Abstract
A method and apparatus for monitoring the capacity of a valve regulated lead acid battery comprising at least one battery monitor connected to the valve regulated lead acid battery; a centralized system connecting the battery monitor through an industry standard data system to a central office; and an alarm connected to the centralized system; wherein, a short-term discharge test is performed on the battery using the battery monitor which provides input parameters for a neural network and fuzzy logic network used in combination with a prediction algorithm to calculate the predicted capacity; and, wherein, the alarm is activated when said predicted capacity falls below eighty percent, when an individual cell voltage is reduced to 1.95 volts or less, or when a system failure occurs
-
Citations
20 Claims
-
1. A method of monitoring the capacity of a valve regulated lead acid battery, comprising:
-
connecting at least one battery monitor to said valve regulated lead acid battery;
connecting said monitor to a centralized system through an industry standard data system that collects and transfers data from the field to a central office;
providing said centralized system with an alarm;
performing a short-term discharge test on said battery using said battery monitor; and
obtaining a predicted capacity using a neural network and fuzzy logic network in combination with a prediction algorithm;
training said neural network for a specific battery type in a lab by determining the actual capacity of a battery, inputting parameters with the actual capacity used as a target, and programming the set of coefficients yielded by said neural network in said battery monitor;
wherein, said discharge test provides input parameters for said networks which in combination with said prediction algorithm calculate said predicted capacity, and wherein, said alarm is activated when said predicted capacity falls below a predetermined percentage, when an individual cell voltage is reduced to about 1.95 volts or less, or when a system failure occurs. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14)
-
-
15. An apparatus for monitoring the capacity of a valve regulated lead acid battery, comprising:
-
at least one battery monitor connected to said valve regulated lead acid battery, said battery monitor having a neural network trained in a lab for a specific battery type by determining the actual capacity of a battery, inputting parameters with the actual capacity used as a target, and programmed with a set of coefficients yielded by said neural network and a fuzzy logic network;
a prediction algorithm, said prediction algorithm using data produced by said neural network and fuzzy logic network, a centralized system connecting said battery monitor through an industry standard data system to a central office;
an alarm connected to said centralized system. - View Dependent Claims (16, 17, 18, 19)
-
-
20. A method of monitoring the capacity of a valve regulated lead acid battery, comprising:
-
connecting at least one battery monitor to said valve regulated lead acid battery;
connecting said monitor to a centralized system through an industry standard data system that collects and transfers data from the field to a central office;
providing said centralized system with an alarm;
performing a short-term discharge test on said battery; and
obtaining a predicted capacity using a neural network and fuzzy logic network in combination with a prediction algorithm;
wherein, said discharge test comprises a four hour discharge of said battery, performed at a discharge rate calculated from the amp-hour size of the battery and a 24-hour period, during which specific data parameters consisting of cell age, open circuit voltage, voltage after one hour of discharge, voltage after three hours of discharge, and voltage after four hours of discharge are logged and three additional parameters including three hour slope, the delta between the three and four hour voltages, and the proximity to two volts of the four hour voltage are determined; and
wherein said neural network is trained in a lab by determining the actual capacity of a battery using the following formula;
(time to discharge to 1.75V×
24 hour discharge rate)/(24×
24-hour discharge rate in amps),inputting said parameters with the actual capacity used as a target, and programming the set of coefficients yielded by said neural network into an EEPROM which is inserted in said battery monitor; and
wherein, said fuzzy logic network comprises dividing total battery capacity into capacity spans of ten percent;
associating a voltage range with each of said ten percent spans, wherein said voltage range was previously determined from said 24 hour rate discharge lab tests;
determining within which of said capacity spans said voltage ranges fall by comparing each of said open, one, three, and four hour cell voltages with the known base lines for the specific battery type;
determining the strength value of the cell'"'"'s voltage within a particular span by the following steps;
calculating the average voltage for each capacity range using the maximum and minimum voltage values;
dividing the cell test voltage (open, one, three, or four hour) by said average voltage;
subtracting the quotient of previous step from one and taking the absolute value of said answer;
performing this calculation for each of cell test voltages (open one, three, and four hour);
adding the calculated values for each of said cells test voltages and dividing the sum by seventy; and
subtracting the quotient of the previous step from one;
determining the capacity range with the most positive strength value;
adjusting the range that is immediately above said capacity range with most positive strength value by adding the strength value of said range to the lowest capacity value of said range to yield the upper capacity limit; and
adjusting the range that is immediately below said capacity range with most positive strength value by subtracting the strength value of said range from the highest capacity value of said range to yield the lower capacity limit;
wherein, said difference between said upper capacity limit and said lower capacity limit yields the fuzzy logic capacity range span; and
wherein, said neural network comprises inputting parameters including open, one-hour, three-hour, and four-hour cell voltages and the age of said battery;
obtaining three more data points from said parameters including slope of the discharge curve, the delta between voltages at three and four hours, and the proximity to two volts of the four hour voltage; and
outputting a number between zero and one;
wherein, said neural network performs calculations as any standard neural net using the coefficients determined from said training of said network using lab data; and
wherein, the output of said neural network is multiplied by the span of the capacity range obtained from the fuzzy logic network and adding this product to said lower capacity limit in order to determine the final capacity prediction.
-
Specification