The Research On Data Acquisition And Processing Of The Air Pump Performance Testing

Air pump is an universal engine, and has been broadly applied to national economy such as oil, chemical industry, electric power, metallurgy, coal, environment protection, shipbuilding, foodstuff curative, city planning and irrigation works etc. And with the development of the technology, the application fields of air pumps have been rapidly improved. The pump industry all over the world has made way for large-scale, high-speed, mechanics-electron integrating, high tech, high reliability, long life span, and autocontrol. Currently, the consumers consider the reliability, life span and auto control of air pumps as the main factors. For this reason, it will be necessary to research high accuracy and efficiency on the test performance system of the air pump. With the development and applications of electronic technology focusing on the computer, the present test system has run to "with sensors, signal modulation circuit and computer-central data collecting system". This system has not only been applied to science experimentation and engineer testing, but also applied to air pumps fields. The system is composed of air pump test bed, sensors, signal modulation circuit, data acquisition card and computer processing. It can acquire dynamic data rapidly, and store, analyze, and compute the test data, then offer the result of the data in time. So, it has improved efficiency of testing and the precision of measurements. In this paper, the air pump data collecting system comprises to the air pump testing bed, computer, and various kinds of parameters sensor, signal modulation circuit, data acquisition card, and control software. It can make the discharge pressure closed loop control; automatically measure main parameters, display data and print. The DC motor drives air pump testing bed, the gas that the air pump compresses is sent into the gasholder through the buffer, and then enters the atmosphere by regulating control wave. According to national standard, we calculate the performance index of the air pump such as shaft power and air discharging capacity through measuring rotational speed, the torque, temperature, pressure, etc. The main research work of this text is as follows: 1. The main task of the data collecting system is to transform the analog signals that the sensors are outputted into the digital signals that the computer can identify, and the computer carries on corresponding calculation and data processing according to different needs. Meanwhile, store and display the data in order to realize to the monitoring of physical quantity. In this paper, we use PCI2000 DAQ of Beijing Art Science & Business Company; it has 12A/D, D/A convector, double end 16 analog quantity input channel, 32output channel, and has the characteristics of high precision, good stability. 2. The sample of the microcomputer control system is a technology that it acquires the state of the matter through indirect method, for example, temperature, pressure, flow, by certain technology transforms them into electronic signals, and then changes the electric signals into the digital signals. Because of the objective reason or the switch technology, the sampling data is mixed by a small amount of noise data in numerous conversions, has influenced the accuracy of the final data. In order to prevent the impact on the result of the data, besides adopting more scientific sampling technology, we will adopt some essential technological means to deal with the initial data; digital filtering is the most basic method. In this paper, we aim at PID control algorithms, adopt first-order filtering algorithm. The existence of different interferences causes the deviation e (n)instable in the measuring channels, thus influence the precision of the control algorithm, and enable controlling systematic function to drop entirely. For this reason, we consider utilizing digital filtering to deal with the interference in the measuring channels, so as to ensure the deviation stability. 3. There are a large number of experimental data during the measurement, through curve fitting and regress analysis to deal with these seemingly irregular data, on terms that sure (if the errors allow), make the data close to the best function expression, so during the course of data processing and automation, it cannot only reduce the cost of the hardware collocation but also strengthen the real-time character of the system controls, and improve the rate and accuracy of measurement. According to a group of two-dimensional data, namely several points on the level, demand to confirm a piece of one-dimensional function curve, make the points best to be close to the curve on the whole, this is the thought that the data fits to the curve, for short, the curve fitting. At present, Most methods of curve fitting are all built on the basis of least square method principle, the basic thought of least square method is: confirming function type at first, setting up error equation through error analysis, on terms that error value is minimum, educe the corresponding regular equation groups (the equation groups of linear algebra), receive the coefficient of regressing analysis, thus set up the fitting multinomial. 4. Combining proportion, integral and differential of deviation is the basic thought of PID control; the three parameters through linear combinations make a controller, and control the target, so the quality key to PID adjusting capacity is the choices of proportion K p, integral K iand differential K d. K pvalue influences systematic response speed and precision; K ivalue influences the steady precision of the system; K dvalue influences the systematic dynamic characteristic, mainly suppress the changes which respond the error. Because of the influence of K p,K i,K don the system is interrelated with each other, and the parameter state of traditional PID adjustor can not change accordingly to the target's state, make the system unable to get the ideal control result. And the genetic algorithm is seeking the excellent algorithm of a kind of simulation natural selection and evolution course, can change according to the target's state, make PID adjustor parameter change thereupon. This thesis combines genetic algorithms and PID controls to reach the satisfactory control result. This kind of method is to encode the control parameters of PID into the chromosome of the genetic algorithms and optimize online them with the genetic algorithms. In this paper, we use the control algorithm that PID combines with genetic algorithm, which has solved controlling precision and relatively lower efficiency...