algorithm Related Articles

Introduction In the FOC(Field Oriented Control) algorithm, the sampling current is the basis of the algorithm implementation and a very important part. So accurate current sampling can bring better result to the algorithm. In other words, if the current sampling is accurate, it will be very helpful for the subsequent coordinate transformation to obtain required results. From this we can see the role of current sampling in the entire FOC algorithm. Understanding Field-Oriented Control Catalog Introduction Ⅰ Current Sampling Method Ⅱ Three Sampling Methods and Precautions 2.1 Single-resistor Sampling 2.2 Dual-resistor Sampling 2.3 Triple-resistor Sampling Ⅲ The Key to Sampling Ⅳ Delay Source Ⅴ Delay Type and Typical Time Ⅵ Analysis in Details 6.1 PWM Dead Time Insertion 6.2 Optocoupler Delay and Pre-Driver Delay 6.3 Transistor Switching Delay 6.4 Other Delays Ⅶ FAQ Ⅰ Current Sampling Method In motor control, the current sampling method is generally to use PWM to trigger ADC to convert. Taking SoC(System-on-a-Chip) as an example, the ADC module will be configured to automatically sample and trigger conversion. When the trigger point set by the PWM module matches, the signal will be given to the ADC module. At this time, the sampling switch in circuit will be disconnected, and then the ADC module will start to convert, and the voltage of the corresponding sampling current can be obtained after the conversion is completed. The AD value of the signal, you can use this value in the program to write and verify the algorithm. Figure 1. Current Sampling Time Ⅱ Three Sampling Methods and Precautions Current sampling is the basis of FOC, including current sensor sampling and resistor sampling. Resistor sampling is widely used for its simple and low-cost characteristics. The method includes single-resistor sampling, dual-resistor sampling, and triple-resistor sampling. 2.1 Single-resistor Sampling The biggest difference between the single-resistor and the other two methods is that it cannot obtain two current signals at the same time. Even if two current signals are obtained, there is an error in estimating the third current signal. The formula Iu+Iv+Iw=0 is conditional, that is, the three currents must be recorded at the same time. When the inductance of the motor is larger, the two currents obtained are closer to the real situation. When the inductance is small, the deviation may be relatively large. So if the inductance of the current is large, single-resistor sampling can be selected.This method requires two samplings in one PWM cycle. In this case, it is necessary to analyze the switch state in the algorithm to clarify which phase current the reconstructed current corresponds to at the time of sampling. 2.2 Dual-resistor Sampling In the case of dual-resistor sampling, the sampled two-phase current must be used directly. Even if there is a deviation, it needs to be used. This method cannot be used to calculate the third-phase current based on the other two-phase sampling like the triple-resistor sampling. That is to say, this method needs to consider the problem of the sampling window. If the sampling current is to be guaranteed to be accurate, the sampling window must be large enough. To make the sampling window large enough, the PWM waveform needs to be deformed. But this will increase the execution time of the algorithm. The advantage of this approach is to reduce a current-sense resistor and an op amp.As shown in the figure below, the front of the red circle is the oscillating area. If the sampling window is small, only the oscillating area will not be able to obtain an accurate current. To process the sampling window, you can refer to the following figure, so that the obtained current will be more accurate. Figure 2. Current Sampling Zone 2.3 Triple-resistor Sampling This method is the simpler among the three methods. It directly uses three current-sensing resistors to sample the three-phase phase current of the motor, and the result obtained in this way is relatively straightforward. Using the formula Iu+Iv+Iw=0, recalculate the phase current of one phase with a small sampling window. So that the accuracy of the result obtained is the highest, and the implementation of the following related algorithms is easier. It is the advantage of this method. However, three current-sense resistors and three op amps are used, the hardware cost will be higher than the other two.   Ⅲ The Key to Sampling The current sampling includes peak current and average current sampling. Generally, the most common is the average current sampling and its control, so there are actually two ways to sample the average current. One is that the current-sense resistor is placed on the upper bridge of the inverter bridge. The other is that the current-sense resistor of the inverter bridge is connected to the lower end of the lower bridge.The general method is the latter. The current detection circuit corresponding to this method is relatively simple, and the corresponding power consumption will also be reduced. In this case, the freewheeling current is collected at the lower end, and then we can sample at the midpoint of the lower bridge opening. At this time, the corresponding current reflects the average current, so the corresponding current control is the average.Then, if we use the three-resistor sampling method, the selected ADC module must have at least the function of simultaneous sampling of three channels. So as to ensure that the three-phase currents obtained by sampling are the currents at the same time, and at this time, to meet the condition, Iu+ Iv+Iw=0.In the case of dual-resistor sampling, there are only two sampling resistors, and the obtained current cannot use the formula Iu+Iv+Iw=0. Therefore, even if the sampling window is small, if the algorithm is not processed, the double-resistor scheme has limitations. In order to get a better adaptation to the scene, algorithm compensation must be performed on the dual-resistor method, which is also the key point of it.Similarly, for the single-resistor sampling way, the corresponding current needs to be obtained according to different switch combinations, and it needs to be sampled twice in a PWM cycle. This method cannot satisfy Iu+Iv+Iw=0, and can only be determined by an algorithm. Compensation and correction are performed, so the single-resistor method is more difficult to take. However, if the difficulty can be solved, this method is the best and cheapest one.   Ⅳ Delay Source During the development of the motor-driven FOC control, have you encountered the situation that the motor is too noisy, inefficient or even unable to operate? All of this may be due to sampling anomalies of the phase currents, resulting in the inability to reconstruct the correct three-phase currents in the FOC algorithm. Here is an analysis of a factor that affects current sampling: the delay source.In the motor drive FOC control of double-resistor sampling, the sampling point is set as the middle moment when the lower tube of the drive bridge is turned on. Note that this is the middle moment when the lower tube of the drive bridge is turned on, not the middle moment of the PWM cycle output by the MCU. There are as many as seven delay sources in this typical drive topology because the PWM is calculated from the MCU to the ADC module where the current signal is sent to the MCU. Figure 3. MCU Output Ⅴ Delay Type and Typical Time The table below details the seven sources of delay that exist in motor drive system topologies and their typical timings. These delays will be superimposed together, and the effect is that the actual output PWM waveform lags behind the PWM waveform that the MCU calculates the expected output. According to this calculation, the phase current sampling point needs to lag the middle moment of the MCU calculating the expected output PWM waveform. Delay Type Typical Time PWM Dead Time Insertion 100ns-2μs Optocoupler Isolation to Pre-driver 40ns-300ns Pre-driver Switch Delay About 50ns MOSFET Switching Time 100ns-1μs Amplifier Delay <1μs Low-pass Filter Delay 1-2μs ADC Delay 50ns-200ns   Ⅵ Analysis in Details 6.1 PWM Dead Time Insertion In the three-phase brushless motor drive system, three bridge arms are required to control the current flow of the phase line, and there are two power devices on each bridge arm, such as MOSFET and IGBT. The pair of power devices cannot be turned on at the same time, otherwise a short circuit will occur. Here MOSFET is used as a power device to illustrate. In the control, dead time must be inserted to ensure that the upper and lower MOSFETs are not turned on at the same time. Typical values of dead time may be between 100ns and 2μs, depending on various factors in the system, such as MOSFET drive voltage and type.After the required PWM waveform is inserted into the dead time, what you get is that both the PWM midpoint and the rising edge are shifted to the right. When using the FOC control algorithm calculates the proper PWM, we start seeing the first delay, recording the dead time. Figure 4. Dead Time Insertion 6.2 Optocoupler Delay and Pre-Driver Delay The signal response of the various optocouplers and pre-drivers causes additional delays between the moment the MCU controls the FTM module to output the PWM waveform and the moment the MOSFET gate is controlled. The output of the pre-driver is delayed by a period of time (Delay1) compared to the waveform output from the MCU pins. Figure 5. Delay 1 6.3 Transistor Switching Delay Through the pre-driver, the PWM waveform reaches the MOSFET transistors, but due to their inherent characteristics, all transistors take a certain amount of time to turn on and off. This delay time varies depending on the transistor type and the voltage level required to switch between on/off. Delay 2 is the total delay between the theoretical switching point (CMP2) of the phase line voltage and the instant of the actual switching point. Figure 6. Delay 2 Finally, the gate voltage reaches the level that can make the transistor turn on, the current passes through the phase line and the sampling resistor, and a voltage difference is generated across the sampling resistor. The red waveform is the phase current waveform in an ideal state. At this time, there is a total delay time between the midpoint of the PWM cycle calculated and generated by the MCU, and the "phase current midpoint shift" is shown in the figure. Figure 7. Phase Current Midpoint Shift 6.4 Other Delays As shown in the figure below, the final delay chain that affects the current sampling is formed by the amplifier slew rate, the low-pass filter on the MCU pins, and the ADC slew rate. The time marked by the red circle in the figure is the correct current sampling time. It can be seen that the phase current sampling point is greatly delayed compared with the PWM midpoint output by the FTM. Figure 8. Other Delay In all and electrical and electronic circuits, there will be signal delay problems. And it is impossible to completely eliminate them, but the impact can be reduced by selecting low-delay devices. In the motor drive, in addition to selecting the appropriate device, it is also necessary to perform software compensation for the signal delay. The precise delay time of these delay sources mentioned in the article can be obtained by oscilloscope and calculation, and the correct current sampling time can be obtained by compensating for these delays in software. In this way, the data collected at the correct moment can be used as the data source for reconstructing the three-phase current of the motor in the FOC control.   Ⅶ FAQ 1. What is FOC algorithm?Field-oriented control (FOC), or vector control, is a technique for variable frequency control of the stator in a three phase AC induction motor. 2. What is FOC drive?Vector control, also called field-oriented control (FOC), is a variable-frequency drive (VFD) control method in which the stator currents of a three-phase AC or brushless DC electric motor are identified as two orthogonal components that can be visualized with a vector. 3. What is FOC brushless motor?FOC implementation allows the BLDC motor to run more efficiently (high power factor and better light load efficiency), more smoothly (lower torque ripples) with quick dynamic response (better dynamic performance to load and speed changes). 4. What is FOC in BLDC motor?Field oriented control (FOC) is an important control approach for Brushless DC motors. It resembles sinusoidal commutation but adds a major mathematical twist. Figure 3a shows control schemes for both sinusoidal commutation and field oriented control. 5. How is Bldc phase current measured?With a BLDC motor use an ac voltmeter to measure the voltage between any 2 wires of the 3 motor wires and then convert the line-to-line voltage to the phase voltage value by dividing the line-to-line voltage by 3 =1.73. 6. Do BLDC motors have inrush current?Handle Peak Inrush Current of a BLDC Motor to protect the Power Supply. Summary: BLDC motors have a Peak current on startup which is 3x or more the rated current. The motor has a rated current of 7.3A. 7. What causes motor inrush current?When an electrical device, such as an AC induction motor, is switched on, it experiences a very high, momentary surge of current, referred to as inrush current. ...The interaction of these two magnetic fields produces torque and causes the motor to turn.

（Researchers have developed an algorithm that allows residential customers to share power from the renewable energy sources in their homes during an outage.） If you think you can use the solar panels on your roof to power your home during an outage, think again. During an outage, while your home remains connected to the grid, the devices that manage your solar panels are powered down for safety reasons. In other words, this permanent connection to the grid makes it impossible for homeowners to draw on power generated by their own renewable energy resources. A team of engineers at the University of California San Diego wants to change this. They have developed algorithms that would allow homes to use and share power from their renewable energy sources during outages by strategically disconnecting these devices, called solar inverters, from the grid. The algorithms work with existing technology and would improve systems' reliability by 25 to 35 percent. Researchers detail the algorithms and their applications in a paper they presented at the American Control Conference in Seattle, Wash. "We were inspired to start investigating a way to use renewable power during outages after Hurricane Sandy affected eight million people on the East Coast and left some without power for up to two weeks," said Abdulelah H. Habib, a Ph.D. candidate in mechanical engineering at UC San Diego and the paper's first author. Our Society is Dependent upon ElectricityJust a few hours without power can cause massive losses to both product and revenue.We rely on electricity much more than we realize. Even if you live "off the grid," as I did for years, you are still living in a world and a society that is deeply dependent upon electricity. If the power is out for a few hours, we have all experienced that; of course you'll be fine. Maybe you will be a little bored and inconvenienced, but if the outage is lengthy and widespread, the consequences can be much more severe, even deadly. What would happen if the electricity was out for a week?Every year, 7 million customers experience power outages. Outages that last more than 5 to 10 minutes cost customers more than $80 billion each year. How the Algorithm WorksThe innovation here is the algorithm's capability to prioritize distribution of power from renewable resources during an outage. The equations take into account forecasts for solar and wind power generation as well as how much energy storage is available, including electric vehicles, batteries and so on. The algorithm combines that information with the amount of energy that the residents are projected to use as well as the amount of energy that a cluster of homes can generate.The algorithm could also be programmed to include a priority function, based on different parameters. For example, customers who are willing to pay more could get priority to get power during an outage. Or customers who generate more energy than they produce during normal operations would not lose power during an outage. More importantly, the algorithm could give priority to customers who are in urgent need of power, because they use life support equipment, for example. Ref.KY605-LC-R064R5PKY605-0860-0004