Part 1: Introduction to Testing Battery Management System (BMS) Software

Summary

In this video, MathWorks' engineers Francesco Alderisio and Maurizio Dalbard introduce methods and techniques for verifying, validating, and testing Battery Management System (BMS) requirements using Simulink. They demonstrate the model-based design process, emphasizing the importance of early validation to catch errors efficiently. The session covers the system analysis, including the plant components and the BMS ECU, and highlights the role of Stateflow in implementing state machines for different operational modes. The presentation also touches on the Model-Based Design workflow's benefits, including its ability to meet stringent certification standards like ISO 26262 for automotive software development.

Takeaways

• 🔧 Francesco Alderisio and Maurizio Dalbard from MathWorks are presenting methods and techniques for verifying, validating, and testing BMS requirements using Simulink before software deployment.
• 🛠️ The system analysis includes a controller, plant components like a battery pack, recharge circuit, and charger/load, with physical connections and electrical components highlighted.
• 🔌 The pre-charge circuit has six switching devices that need to be controlled to prevent electrical spikes.
• 🔋 The battery pack model includes six cells in series and thermal behavior via convection, along with a cell monitoring unit for charge balancing.
• 🤖 The BMS ECU model comprises different components for tasks like balancing logic, SOC estimation, and a state machine implemented in Stateflow.
• 🔄 The state machine includes parallel states for standby, charging, driving modes, fault monitoring, and contactor control to prevent spikes.
• 🔍 The Model-Based Design workflow is emphasized for early identification of errors, with a shift from ambiguous designs to graphical models for precise software specification.
• 📈 The cost of finding bugs increases over time, making early validation in the design process crucial for cost-effectiveness.
• 🛑 Simulation is an early validation tool, but a more rigorous approach is needed for systematic testing to ensure all requirements are met.
• 📚 The complete Model-Based Design process allows for earlier and more cost-effective error identification, with simulation and testing integrated into the workflow.
• 🏢 Companies like LG have used this workflow to develop BMS software for hybrid vehicles, achieving ISO 26262 SLC certification for AUTOSAR code.
• 📘 The Model-Based Design workflow is part of a development process approved by TUV certification authority, with documentation included in the IEC Certification Kit and DO Qualification Kit.

Q & A

What is the role of Francesco Alderisio in the presentation?

Francesco Alderisio is an application engineer at MathWorks, and he is the presenter who introduces the topic of verifying, validating, and testing BMS (Battery Management System) requirements using Simulink.

Who is Maurizio Dalbard and what does he contribute to the presentation?

Maurizio Dalbard is a Senior Application Engineer at MathWorks. He provides an overview of the Model-Based Design workflow, focusing on verification and validation, and discusses the importance of systematic testing in the development process.

What is the purpose of the system analysis presented in the script?

The system analysis is aimed at understanding the components and connections within a BMS, including the controller, plant, battery pack, recharge circuit, charger, and load, to ensure proper control and avoid undesired electrical spikes.

What are the different components of the plant mentioned in the script?

The plant consists of a battery pack, a recharge circuit, and a charger and load setup, each with specific components such as switching devices and thermal behavior models.

How many cells are connected in series in the battery pack model?

In the battery pack model, there are six different cells connected in series.

What is the role of the cell monitoring unit in the BMS?

The cell monitoring unit is responsible for controlling other switching devices to maintain a balanced state of charge across the different cells in the battery pack.

What are the four different model references mentioned in the BMS ECU?

The four model references in the BMS ECU include a balancing logic, an SOC (State of Charge) estimation model, a state machine implemented in Stateflow, and fault monitoring states.

What is the significance of the state machine in the BMS ECU?

The state machine in the BMS ECU is crucial as it implements logic to manage different states such as standby, charging, driving modes, and fault modes, ensuring the system operates correctly under various conditions.

What are the advantages of using Simulink for verifying and validating BMS requirements?

Simulink allows for the use of graphical models or executable specifications that provide precise meanings, making it easier to identify errors early in the design phase, reducing the cost and complexity of bug detection.

How does the Model-Based Design process help in reducing the cost of finding bugs?

The Model-Based Design process enables early validation of the design through simulation, which is more cost-effective than testing in the final phase, as errors are cheaper and easier to identify early on.

What is the significance of the IEC Certification Kit and DO Qualification Kit mentioned in the script?

The IEC Certification Kit and DO Qualification Kit provide complete documentation for the Model-Based Design workflow, which is essential for developing critical embedded software that meets the standards required for certification by authorities such as TUV.

Outlines

00:00

🛠️ Introduction to BMS Verification and Testing in Simulink

Francesco Alderisio and Maurizio Dalbard, both from MathWorks, introduce a session focused on Battery Management System (BMS) requirements verification, validation, and testing using Simulink. They discuss the importance of adopting methods and techniques to ensure the software's readiness for deployment on embedded microprocessors. The session begins with an analysis of the system components, including a controller, plant, battery pack, recharge circuit, and charger with their respective connections and functions. The battery pack is further detailed with its series-connected cells and thermal behavior modeling. The BMS ECU is highlighted with its various model references, including balancing logic, SOC estimation, and a state machine implemented in Stateflow, which manages different states such as standby, charging, driving, and fault monitoring. The overview also touches on the control of contactors to prevent electrical spikes. The presenters encourage viewers to refer to another webinar for more details on the passive balancing logic.

05:02

🔍 Model-Based Design for BMS Software Verification and Validation

Maurizio Dalbard provides an in-depth look at the Model-Based Design workflow, emphasizing the verification and validation stages in the development of critical embedded software like a BMS. He explains the transition from requirements to design specifications and the challenges of identifying errors in traditional processes due to the lack of simulation capabilities. Dalbard highlights the benefits of using Simulink for creating graphical models or executable specifications that facilitate early and cost-effective error detection. The Model-Based Design process is positioned as a way to validate the design early in the development cycle, reducing the cost of bug fixes over time. The session outlines the need for a more rigorous approach to testing the model systematically to ensure all requirements are met and the system functions correctly. Dalbard also mentions the integration of verification, validation, and test activities into the workflow, including component and system testing, review, and static analysis. The presentation concludes with examples of companies like LG, which have successfully used the Model-Based Design workflow to achieve ISO 26262 SLC certification for AUTOSAR code in BMS development, with complete documentation provided in the IEC Certification Kit and DO Qualification Kit.

Keywords

💡Application Engineer

An application engineer is a professional who applies engineering principles to design, develop, and test products or systems. In the context of the video, Francesco Alderisio and Maurizio Dalbard are application engineers at MathWorks, a company known for its software for engineers and scientists. They are demonstrating the use of Simulink, a graphical programming environment, to verify and validate battery management system (BMS) requirements.

💡Simulink

Simulink is a block diagram-based environment for model-based design of dynamic systems such as control systems associated with embedded systems. It is used in the video to show how BMS requirements can be verified and validated before deploying the software onto an embedded microprocessor. The script mentions using Simulink to replace ambiguous designs with graphical models that can be simulated and tested.

💡BMS Requirements

BMS stands for Battery Management System, which is a critical component in electric and hybrid vehicles to monitor and control battery charging and discharging. In the video, the term 'BMS requirements' refers to the specific needs and specifications that the BMS must meet. The script discusses methods to verify and validate these requirements using Simulink before actual deployment.

💡Embedded Microprocessor

An embedded microprocessor is a microprocessor specifically designed for use in embedded systems, which are systems designed for a particular function within a larger mechanical or electrical system. In the video, the goal is to deploy the software that has been verified and validated in Simulink onto an embedded microprocessor, which will control the BMS.

💡Plant

In the context of control systems, a 'plant' refers to the physical system or process that is being controlled. In the video, the 'plant' includes components like the battery pack, recharge circuit, charger, and load, which are part of the system that the BMS is designed to manage.

💡Switching Devices

Switching devices are components that can open or close circuits, controlling the flow of electricity. In the script, they are mentioned in the context of the pre-charge circuit and cell monitoring unit, where they need to be controlled to avoid undesired electrical spikes and to balance the charge of the battery cells.

💡State Machine

A state machine is a computational model that represents a system's behavior as a sequence of states and transitions between those states based on inputs. In the video, the state machine is a key component of the BMS ECU, implemented in Stateflow, which includes different states like standby, charging, and driving modes.

💡Fault Monitoring

Fault monitoring is the process of detecting and responding to faults or errors in a system. In the video, it is part of the state machine where faults in current, temperature, or cell voltages are monitored to ensure the safe operation of the BMS.

💡Model-Based Design

Model-Based Design is a design approach that uses graphical models or executable specifications to design systems. It is highlighted in the video as a process that helps to identify errors early and reduce costs by allowing for simulation and testing before coding. The video script discusses how Simulink facilitates this workflow.

💡Verification and Validation

Verification and validation are processes used to ensure that a system meets its requirements and performs as expected. In the video, these processes are crucial for testing the BMS model in Simulink to gain confidence in its correctness before deployment. The script emphasizes the importance of a rigorous approach to systematically test the model.

💡IEC Certification Kit

The IEC (International Electrotechnical Commission) Certification Kit is a set of tools and documentation that helps in achieving compliance with international standards for electrical, electronic, and related technologies. In the video, it is mentioned that the Model-Based Design workflow used by LG for BMS software development is reviewed and approved by TUV, a certification authority, and includes documentation according to IEC standards.

Highlights

Francesco Alderisio and Maurizio Dalbard from MathWorks demonstrate methods and techniques in Simulink for verifying, validating, and testing BMS requirements before deploying software onto an embedded microprocessor.

The system being analyzed consists of a controller, plant with a battery pack, recharge circuit, charger, and load, highlighting the physical connections between electrical components.

Six switching devices in the pre-charge circuit need to be controlled to avoid undesired spikes between the battery, charger, and load.

The battery pack model includes six cells connected in series and thermal behavior modeled by convection.

The cell monitoring unit allows for control of switching devices to maintain a balanced state of charge across different cells.

The BMS ECU model references include balancing logic, SOC estimation, and a state machine implemented in Stateflow with different parallel states for various modes and fault monitoring.

The Model-Based Design workflow is presented with a focus on verification and validation, emphasizing the importance of catching errors early in the design phase.

Simulink replaces ambiguous designs with graphical models or executable specifications to provide precise meanings and facilitate code generation.

Model-Based Design allows for earlier validation of the design where errors are cheaper and easier to identify.

Simulation is a great way to validate behavior early, but a more rigorous approach is needed for systematic testing to ensure all requirements are met.

The complete Model-Based Design workflow enables developers to gain confidence in the correctness of their designs through verification, validation, and testing activities.

The Model-Based Design workflow for developing critical embedded software is reviewed and approved by the TUV certification authority, with complete documentation included in the IEC Certification Kit and DO Qualification Kit.

LG used the Model-Based Design workflow to develop BMS software for a hybrid vehicle, achieving ISO 26262 SLC certification for AUTOSAR code.

The webinar provides an overview of the model and references to more detailed information on passive balancing logic and other components of the BMS development process.

Maurizio Dalbard emphasizes the importance of a rigorous testing approach in the Model-Based Design process to ensure the correctness and functionality of the design.

The presented workflow and techniques are part of a development process that has been certified and documented according to industry standards, ensuring reliability and compliance.

Transcripts

Hello everyone, my name is Francesco Alderisio and I'm an application engineer at

MathWorks. Today I've been joined by my colleague Maurizio Dalbard, Senior

Application Engineer. Today we will show you the methods and techniques

that you can adopt in Simulink to verify validate and test BMS requirements

before deploying the software onto an embedded microprocessor. So let's start

by analyzing the system that we have here. So we have a controller, and then we

have a plant, so let's go into the plant. And here we can recognize three

different components: battery pack, a recharge circuit, and then charger and

load. And here you can see the physical connections between the electrical

components. So let's go into the pre-charge circuit and here you can see

that we have six different switching devices that we need to appropriately

control so that we can avoid undesired spikes between the battery and between

the charger and the load. Let's go back and now let's analyze the battery pack.

So here we have two more components in particular. Here we can see that we have

six different cells connected in series, and then we have also modeled the

thermal behavior by convection. Okay then here we have the cell monitoring unit. So

in particular, here we have other switching devices that we can

appropriately open and close with our controller, so as the balanced state of

charge of the different cells.

Okay so now let's go into the BMS ECU. So here we have four different model

references. So we have different components that do different tasks. In

particular we have a balancing logic to balance the set of charge of the cells.

We have an SOC estimation model to estimate the set of charge of our cells

and then we have here the main component which we will analyze today, which is the

state machine. And here we have implemented a logic in Stateflow. In

particular we have different states that run in parallel, so we have the main

state machine in which we can see that we have a standby mode charging mode a

driving mode and so on. And we also have a fault mode. Then we have a fault

monitoring state, where we can monitor folds on the current faults on a

temperature or folds on the cell voltages. And then we have two more

parallel states where we can actually control the contactors in our model so

that we can avoid spikes. We have one for the inverter and then one for the

charger. Okay so this was a very brief overview of the model. If you want to

have more details on how this was built and more details on the passive

balancing logic please refer to the webinar battery management system

development Simulink where our colleague Javier will walk you through it.

Thank you Francesco for the introduction. Hi everyone, my name is Maurizio Dalbard. I'm in

application engineering at MathWorks. Before starting by seeing in practice

the methods and techniques you can adopt in Simulink to verify, validate, and test, I

would like to give you another view of the Model-Based Design workflow with a

more specific focus on verification and validation in the traditional process. We

start from the requirements that are refined into design specification. And

under is a gap in the process that we have to fill to the end written code.

Using this process, it is difficult to find the errors especially in the

initial design, because we can't simulate and we can just test in the final phase.

Many studies also show that the cost of finding a bug increases over time. In

the Model-Based Design process, Simulink replaces ambiguous designs using

graphical models or executable specifications that have precise

meanings, adding some some information about the software. And the targets will

obtain the model ready for cogeneration so that similar designs can be

automatically converted into code that we can download in our final target. With

the complete Model-Based Design process, you can validate your design here

earlier where errors are cheaper and easier to identify. Simulation

is a great way to validate behavior early but it is mainly among our forces

and largely ad hoc. And to define if all requirements are met and functioning

correctly we need to take a more rigorous approach to systematically test

our model in order to get the complete confidence in the correctness

of our designer using a complete Model-Based Design workflow.

So now adding some verification validation and test activities we have

to do some component and system testing, review and static analysis equivalent

testing, and we run and checked in, for example the motor position workflow is

complete. Step by step we'll touch some of the many techniques we can leverage

in Simulink to get more confidence in the correctness of our designs. This workflow is

a workflow to develop critical embedded software that was

reviewed and approved by the TUV certification authority. And there is

complete documentation describing it included in the IEC Certification Kit

and DO Qualification Kit according to the standards. In this slide you can see

LG, using the Model-Based Design workflow, developed software for a BMS

for a hybrid vehicle achieving the ISO 26262 SLC certification

for AUTOSAR code.