1. 定义一个字典，将每个州和地区的获胜者以及他们的票数存储在其中。
2. 对于每年和每个州，遍历所有地区的选举结果并计算获胜者的得票率。
3. 如果得票率高于55％（5%优势），则将获胜者及其票数添加到字典中。
4. 最后，输出字典中的所有获胜者及其得票数。

示例代码：

election_results = {
    2020: {
        "California": {
            "District 1": {
                "Candidate A": 10000,
                "Candidate B": 8000
            },
            "District 2": {
                "Candidate A": 12000,
                "Candidate B": 10000
            }
        },
        "Texas": {
            "District 1": {
                "Candidate C": 5000,
                "Candidate D": 4000
            },
            "District 2": {
                "Candidate D": 6000,
                "Candidate E": 5500
            }
        }
    }
}

winners = {}

for year in election_results:
    for state in election_results[year]:
        for district in election_results[year][state]:
            candidate, votes = max(election_results[year][state][district].items(), key=lambda x:x[1])
            total_votes = sum(election_results[year][state][district].values())
            if (votes/total_votes) > 0.55:
                if year not in winners:
                    winners[year] = {}
                if state not in winners[year]:
                    winners[year][state] = {}
                winners[year][state][district] = {candidate: votes}

print(winners)