Modular parallel operation of inverters can greatly improve system flexibility, breaking the limitations of inverters in terms of power level. Users can combine system power according to their needs, and it is also easy to implement redundant design, thus having the advantages of high reliability and ease of high-power scaling. Research on parallel inverter communication monitoring technology is an issue that must be solved in the process of AC power systems evolving from traditional centralized power supply to distributed power supply and even intelligent power supply system power supply modes. This paper introduces a parallel inverter communication monitoring system based on CAN fieldbus. The system makes full use of the internal resources of TI's TMS320LF2407A DSP chip, acquires and parses field control data from each parallel module through the CAN bus, responds to field operations with strong real-time requirements, and realizes scheduling and monitoring of module operation. It has the advantages of simple structure, convenient expansion and high reliability.
1 System Composition
1.1 System Network Structure
The system composition is shown in Figure 1. The system consists of a monitoring host, parallel power supply modules, and a CAN bus. The inverter power supply module mainly comprises a PWM inverter, a DSP controller, signal sampling and load current sharing, and communication control components, using a TMS320LF2407A (DSP) as the control core. The DSP controls the output of PWM pulses based on the difference between the feedback voltage and current values and the set values, driving the inverter bridge power switch to output a sinusoidal voltage with adjustable frequency, amplitude, and phase. The LF2407A's embedded CAN controller receives commands from the CAN bus to control the parallel inverter power supply module.
This topology is characterized by multiple network communication nodes sharing a single transmission line, resulting in high channel utilization, simple connections, low cost, and high system reliability. Information transmission uses the CAN communication protocol, with twisted-pair cable as the transmission medium. To further enhance the system's anti-interference capabilities, opto-isolation measures can be added between the controller and the transmission medium.
1.2 Main System Hardware
The hardware module circuit is shown in Figure 2. Because the LF2407A chip itself contains an embedded CAN controller, the hardware design is relatively simple. Only a transceiver SN65HVD232D needs to be added to realize the interface between this node and the bus.
The SN65HVD232D is an interface chip between TI's CAN protocol controller and physical bus, conforming to the ISO11898 standard. It provides differential transmit capability to the bus and differential receive capability to the CAN controller. A 120Ω matching resistor is added at the termination to ensure interference immunity and reliability of data communication.
2 Software Design
2.1 Programming for Baud Rate Setting
The transmission rate of the CAN bus is related to the maximum distance between two nodes. Table 1 shows the relationship between the maximum distance between any two nodes in a CAN bus system and the bit rate. The table also provides the values of the bit timer configuration registers BCR2 and BCR1 of the LF2407A. These values are related to the master clock frequency of the LF2407A. Generally, the bit rate can be calculated using the following formula:
Baud rate = ICLK / [(BRP+1)×bitTime] (1)
The identifier serves as the name of the message and is first sent to the bus during the arbitration process. It is used in the receiver's acceptance testing and in determining access priority during the arbitration process.
The Remote Transmit Request (RTR) bit is used to determine whether to transmit a data frame or a remote frame. When the RTR is high, the CAN controller transmits a remote frame; when it is low, it transmits a data frame.
The Data Length Code (DLC) is used to determine the number of data bytes to be sent in each frame, with a maximum of 8 bytes.
The control command indicates the meaning of this frame. The meaning of the control command words in this article is shown in Table 2.
2.3 Program Flow Design
Master-slave control is a relatively mature method for parallel control of inverter power supplies. This design adopts a contention-for-master master-slave control strategy to achieve communication monitoring of the parallel inverter power supply system.
If we set email address 3 as the sending email address for non-broadcast messages, email address 2 as the receiving email address for non-broadcast messages, email address 4 as the sending email address for broadcast messages, and email address 0 as the receiving email address for broadcast messages, then the flowcharts for the message receiving and sending procedures are shown in Figure 4 and Figure 5, respectively.
As shown in the flowchart, the entire communication system mainly consists of a master node and multiple child nodes. Each child node uses mailbox 4 to periodically broadcast master contention requests to the bus at regular intervals to detect whether the master node is functioning correctly. If it is functioning correctly, the master node will send a response opposing the master contention request, which is also broadcast on the bus. Therefore, every node on the network receives both the master contention request and the response opposing the master contention request. Mailbox 0's function is to receive and distinguish between these two types of information, determine its own status, and decide whether to use mailbox 4 to send a response opposing the master contention request. Mailbox 2's main function is to receive control information from the master node and notify mailbox 3 to send its response.
Figure 5 shows the information transmission process in detail. The program listing for the CAN communication mastering part of this system, written using DSP assembly instructions, is as follows.
LDP #DP_CAN
SPLK #0040H, TCR; Mailbox 4 sends a leader contention request LDP #DP_PF2
LAR AR7, #4H
LDP #DP_CAN
MAR *, AR4; Number of sends: LAR AR4, #0FFFFH
W_TA5: LDP #6
SBRK #01H
SAR AR4, 30H
LACL 30H
BCND W_TA7, EQ
LDP #DP_CAN
BIT TCR, 1; Waiting for response BCND W_TA5, NTC
SPLK #4000H, TCR
LDP #DP_CAN
MAR *, AR4; Number of sends: LAR AR4, #0FFFFH
W_TA7: LDP #DP_CAN
MAR*, AR4; Number of sends: LAR AR4, #0FFFFH
W_TA6: LDP #6
SBRK #01H
SAR AR4, 30H
LACL 30H
BCND W_TA9, EQ; Write email content, configuration parameters given by LDP #DP_USER
BIT CAN_FLAG1, BIT0; Check if a response opposing the master node application has been received BCND W_TA6, NTC; If no opposition is received, modify ID1 and ID0 of mailbox 4. A value of 10 indicates that this node is the master node. LDP #DP_USER
SPLK #00H, CAN_FLAG1
LDP #DP_CAN
SPLK #4000H, TCR; Clear TA4 and MIF4
Call LOOP11; Email 3 sends data. Using fieldbus control technology, a distributed inverter power supply local area control network can be easily constructed, giving the power supply system the characteristics of a field network control system (FCS). This method not only inherits the advantages of distributed control systems (DCS) but also integrates digital communication and intelligent network control. The system introduced in this paper not only simplifies the parallel connection of inverters but also provides stable and reliable data communication for each module. It realizes control functions such as system parameter setting, static current sharing of parallel inverter modules, and module master contention. Furthermore, the system has a simple structure and reliable operation.
