Diesel engine testing is the final critical step in the diesel engine production process, playing a vital role in the manufacturing quality of diesel engines. Its main function is to determine the pass/fail status of the diesel engine by recording and analyzing process parameters under specific testing conditions, and to promptly identify and eliminate faults. 1. Introduction Many currently used diesel engine testing platforms analyze and judge the operating status of a diesel engine during testing based on instrument readings. This is not only inefficient and inaccurate, but also has limited comprehensive analytical and judgment capabilities. To gain a more comprehensive and intuitive understanding of the diesel engine testing process, quickly identify and eliminate potential faults, and improve the analytical and judgment capabilities of testing operators, we developed a diesel engine testing system based on CAN fieldbus, in conjunction with enterprise technological transformation. This system enables simultaneous monitoring and testing of multiple diesel engines. 2. Test System Structure Based on the testing requirements of diesel engines, this system mainly processes various sensor signals during the diesel engine testing process and collects diesel engine operating condition data. The data is then sent to the host computer via the CAN bus, requiring the processing of 16 analog signals and 16 I/O signals. The parameters collected mainly include: oil pressure and temperature, coolant temperature, intake and exhaust temperature, fuel level, starting battery voltage, speed, etc. The key to the diesel engine test system is the introduction of CAN bus technology to form a distributed measurement and control system model based on CAN bus, as shown in Figure 1. Since CAN bus is one of the types of fieldbus, it belongs to the open low-level control network. It is a system applied to the production site to realize bidirectional serial multi-node digital communication between microcomputer measurement and control equipment. Therefore, the distributed measurement and control system based on CAN bus is open, rather than closed and dedicated [1]. This test system distributes the monitoring function to each test bench, and each test bench is monitored by a CAN intelligent node. The composition of each CAN bus node is the same, including: main control unit, CAN bus communication management unit, data acquisition and processing unit, etc. Each node is connected to the host computer through CAN bus and completes communication with each other through the bus. 3. Hardware design of the test system The diesel engine test system adopts a two-level distributed structure. The host computer is a PC, and a PC-CAN bus adapter card is installed in the PCI bus slot of the host computer. In this way, the host and slave computers can be connected together through the CAN bus to form a control network. The lower-level controller uses intelligent nodes composed of an AT89C51 microcontroller and an SJA1000 CAN bus controller. These nodes directly control various field devices (such as sensors, relays, and motors), collect field data, and send data to the CAN bus based on received commands or by actively sending data. Acceptance codes and acceptance mask codes can be pre-set to control which data or commands the intelligent nodes receive from the bus. If certain data requires further complex processing (such as dynamic display), the upper-level computer can receive the data from the bus. When the upper-level computer needs to apply control actions to a specific node, it can communicate with that node point-to-point. When it needs to apply control actions to all nodes simultaneously, it can broadcast commands to the bus. This allows the system to operate normally without the involvement of the upper-level computer, significantly reducing data transmission volume and improving system real-time performance and reliability. The hardware design of the diesel engine testing system mainly involves the PC-CAN adapter card in the upper-level computer and the lower-level CAN intelligent nodes. This section focuses on analyzing the structural composition of the CAN intelligent nodes. In the CAN smart node shown in Figure 2, the core components are the CAN bus controller SJA1000, the CAN bus driver 82C250, and the microcontroller AT89C51. The AT89C51 has two main tasks: first, it is responsible for initializing the CAN controller SJA1000 and controlling the SJA1000 to perform communication tasks such as data reception and transmission; second, it is responsible for acquiring field signals and controlling field devices. The SJA1000 is a CAN controller from Philips, implementing the data link layer and physical layer functions of the CAN bus network. Through programming, a microprocessor can configure its operating mode, control its working state, and perform data transmission and reception, building the application layer on top of it. In this design, to enhance the anti-interference capability of the CAN bus nodes, the SJA1000's opto-isolated CAN bus interface is used. The SJA1000's transmit output TX0 and receive inputs RX0 and RX1 are isolated by a high-speed integrated optocoupler 6N137 and then connected to the TXD and RXD of the CAN bus interface driver chip 82C250. The 82C250 is directly connected to the CAN physical bus. The host computer monitoring software is developed using configuration software. Configuration software, as a user-customizable software platform tool, has developed with the increasing maturity of distributed control systems and computer control technology. Currently, with the gradual promotion of fieldbus technology, fieldbus and open systems have become the external environment on which configuration software grows. This makes it easier for configuration software to connect with a large number of input/output devices, thereby promoting the application of configuration software in fieldbus control systems. By comparing the performance and price of existing configuration software, and in combination with the actual needs of this technical transformation project, the domestic "Century Star" configuration software was selected to develop the monitoring program for the CAN bus system. In order to organically combine the host computer human-machine interface program with the lower computer data acquisition and exchange program, we divided the monitoring program into two parts: applying the server-client structure to the configuration software design of the CAN bus control system, realizing the human-machine interface program as the client-side program and the program that exchanges data with the hardware as the server-side program. (2) Lower computer software: Each diesel engine test stand acts as an intelligent node on the CAN bus, transmitting its detection status and control results to the host computer via the CAN communication interface, and is ready to receive control commands from the host computer at any time. The lower-level control program employs modular programming, including a CAN bus communication management module, a diesel engine operating status monitoring module, an A/D inspection sampling and data transfer module, and an I/O switch signal processing module. Among these, the CAN node communication module is crucial, as it determines the proper functioning of the entire distributed control network. The CAN node communication module consists of a CAN initialization subroutine, a CAN interrupt program, and a CAN data transceiver subroutine, as shown in Figure 4. 5. Conclusion The research content of this paper has been tested and runs well on the test bench of the 135 series diesel engine. Practice shows that the distributed test system based on CAN bus is stable and reliable, and has the advantages of flexible and simple configuration, low cost, high reliability, strong anti-interference ability and good expandability. It can comprehensively monitor the diesel engine testing process, greatly reduce testing time, improve monitoring working conditions, improve the scientific management of equipment, and can detect certain faults early, preventing losses caused by delayed handling.